CN112090478A - Crusher fault diagnosis method based on linear regression - Google Patents

Crusher fault diagnosis method based on linear regression Download PDF

Info

Publication number
CN112090478A
CN112090478A CN202010863644.9A CN202010863644A CN112090478A CN 112090478 A CN112090478 A CN 112090478A CN 202010863644 A CN202010863644 A CN 202010863644A CN 112090478 A CN112090478 A CN 112090478A
Authority
CN
China
Prior art keywords
data
crusher
regression
linear regression
groups
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202010863644.9A
Other languages
Chinese (zh)
Other versions
CN112090478B (en
Inventor
张勇
董星宇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
University of Science and Technology Liaoning USTL
Original Assignee
University of Science and Technology Liaoning USTL
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by University of Science and Technology Liaoning USTL filed Critical University of Science and Technology Liaoning USTL
Priority to CN202010863644.9A priority Critical patent/CN112090478B/en
Publication of CN112090478A publication Critical patent/CN112090478A/en
Application granted granted Critical
Publication of CN112090478B publication Critical patent/CN112090478B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B02—CRUSHING, PULVERISING, OR DISINTEGRATING; PREPARATORY TREATMENT OF GRAIN FOR MILLING
    • B02C—CRUSHING, PULVERISING, OR DISINTEGRATING IN GENERAL; MILLING GRAIN
    • B02C2/00—Crushing or disintegrating by gyratory or cone crushers
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B02—CRUSHING, PULVERISING, OR DISINTEGRATING; PREPARATORY TREATMENT OF GRAIN FOR MILLING
    • B02C—CRUSHING, PULVERISING, OR DISINTEGRATING IN GENERAL; MILLING GRAIN
    • B02C23/00—Auxiliary methods or auxiliary devices or accessories specially adapted for crushing or disintegrating not provided for in preceding groups or not specially adapted to apparatus covered by a single preceding group
    • B02C23/04—Safety devices
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/901—Indexing; Data structures therefor; Storage structures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/11—Complex mathematical operations for solving equations, e.g. nonlinear equations, general mathematical optimization problems
    • G06F17/13—Differential equations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/20—Administration of product repair or maintenance

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Business, Economics & Management (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Human Resources & Organizations (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Operations Research (AREA)
  • Software Systems (AREA)
  • Algebra (AREA)
  • Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Strategic Management (AREA)
  • Development Economics (AREA)
  • Quality & Reliability (AREA)
  • Food Science & Technology (AREA)
  • Marketing (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Educational Administration (AREA)
  • Mechanical Engineering (AREA)
  • Game Theory and Decision Science (AREA)
  • Computing Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Probability & Statistics with Applications (AREA)
  • Disintegrating Or Milling (AREA)

Abstract

A fault diagnosis method based on a linear regression crusher selects host current, transmission shaft amplitude and frequency, fuselage amplitude and frequency of a cone crusher as fault diagnosis parameters, establishes n groups of data recorded after the cone crusher works stably by using a multiple regression method after normalization processing to form a partial regression coefficient vector for diagnosis, performs one-time multiple linear regression on new data by using the n groups to obtain a vector representing the current state, and calculates a correlation coefficient between the current vector and the diagnosis vector so as to reflect the degree of the current operation condition of the cone crusher deviating from the stable state in normal operation; when the correlation degree is below 0.6 but no sharp descending trend appears, the machine is checked after shutdown, the fault type is recorded according to the check result, and the array after multivariate regression is added into a comparison database for next fault diagnosis; when the correlation degree is lower than 0.6 and the rapid descending trend continues to appear, the system should be stopped for inspection as soon as possible, and the fault diagnosis time is saved for enterprises.

Description

Crusher fault diagnosis method based on linear regression
Technical Field
The invention relates to the technical field of technical detection, in particular to a fault diagnosis method for a crusher based on linear regression.
Background
The cone crusher is widely applied to a plurality of departments such as mines, smelting, building materials, roads, water conservancy and the like, and particularly plays a key role in the middle breaking and fine breaking of mine rocks. Due to the fact that the working environment is severe, the failure rate of the cone crusher is high, the maintenance cost is also high, and the failure monitoring and diagnosis are key problems to be solved urgently in mine equipment management.
In order to ensure the enterprise benefit, shutdown detection and maintenance are increasingly not advocated by enterprises. On-line fault diagnosis can meet this requirement, reducing off-line operation time, but also increasing the requirements on detection equipment and processors. Common fault diagnosis methods include neural network algorithms, fuzzy control algorithms, support vector machines, expert systems, and the like. The neural network algorithm and the support vector machine both need training data with classification labels as supports, and particularly the neural network needs a large amount of data supports; the fuzzy control algorithm needs to select a fuzzy membership function by experience, and the establishment of the fuzzy rule is based on a large amount of operation practice; the expert system needs experts to know the fault diagnosis in the field as much as possible to establish a relatively perfect expert knowledge base, which is not suitable for fault diagnosis of the cone crusher.
Disclosure of Invention
In order to solve the technical problems provided by the background art, the invention provides a fault diagnosis method based on a linear regression crusher, which utilizes historical data of a cone crusher during normal operation as training data, and takes the host machine current, the transmission shaft amplitude and frequency, the machine body amplitude and frequency, the lubricating system pressure and the like of the crusher as fault diagnosis parameters for analysis. The method avoids inconvenience caused by the pre-construction of large sample data, ensures the authenticity and reliability of the data, can operate without knowing the relationship between the fault of the crusher and each characteristic quantity in advance, and can avoid missed diagnosis and misdiagnosis caused by the imperfect knowledge base by carrying out fault analysis after the fault is diagnosed. The time for fault diagnosis is saved for enterprises, and the production efficiency of the cone crusher is improved.
In order to achieve the purpose, the invention adopts the following technical scheme:
a fault diagnosis method based on a linear regression crusher selects host current, transmission shaft amplitude and frequency, fuselage amplitude and frequency of a cone crusher as fault diagnosis parameters, establishes n groups of data recorded after the cone crusher works stably by using a multiple regression method after normalization processing to form a partial regression coefficient vector for diagnosis, performs one-time multiple linear regression on new data by using the n groups to obtain a vector representing the current state, and calculates a correlation coefficient between the current vector and the diagnosis vector so as to reflect the degree of the current operation condition of the cone crusher deviating from the stable state in normal operation; when the correlation degree is below 0.6 but no sharp descending trend appears, the machine is checked after shutdown, the fault type is recorded according to the check result, and the array after multivariate regression is added into a comparison database for next fault diagnosis; when the correlation degree is lower than 0.6 and a rapid descending trend continues to appear, the machine should be stopped for checking as soon as possible; the method specifically comprises the following steps:
step one, selecting fault diagnosis parameters and carrying out normalization processing
According to the working characteristics of the cone crusher, the current signal and the vibration signal are rapidly changed when the crusher stably operates, and the cone crusher has the capability of rapidly reflecting the current operation condition of equipment relative to signals such as temperature, oil pressure and the like, so that the host machine current, the transmission shaft amplitude and frequency, the machine body amplitude and frequency, the lubricating oil tank oil supply pressure, the locking pressure, the releasing pressure and the lining plate pressure of the cone crusher are selected as the basis of fault diagnosis;
step two, generating comparison data for fault diagnosis
Acquiring n groups of data after the cone crusher is started and stably operates, and calculating a group of approximate solutions by using a multiple linear regression method to serve as comparison data for fault diagnosis; obtaining n groups of stable operation data through a computer, and performing linear fitting according to a multiple linear regression theory to obtain a vector b for comparison and representing normal state of the crusher;
step three, generating detection data representing the current state of the crusher in a rolling manner
After the cone crusher generates the contrast data for representing the normal state, the rolling updated partial regression coefficient vector b is calculated according to the multiple regression theorynewPerforming linear regression on each n groups of subsequent sampling data, and updating one group of the subsequent sampling data each time to generate real-time updated detection data;
step four, calculating the correlation between the detection data and the comparison data
B, b is calculated according to a correlation coefficient calculation formulanewThe correlation coefficient r between the two is judged according to the threshold value of 0.6 to judge whether the current running state of the crusher is normal, r>0.6 represents that the crusher normally operates, the variation of the correlation coefficient of the cone crusher in stable operation is small, and when the correlation coefficient is reduced to be below 0.6, the cone crusher is in a state which is seriously deviated from the normal operation and needs to be overhauled after shutdown;
step five, obtaining a diagnosis result and taking the diagnosis result as the basis of the next diagnosis
For the crusher diagnosed with the abnormal state, after expert identification, fault types are recorded, a group of partial regression coefficient vectors after multiple regression generated by real-time data when the correlation coefficient is lower than 0.6 is reserved, the partial regression coefficient vectors are used for judging fault types when faults occur next time, if the partial regression coefficient vectors are not the same type of faults, namely the correlation coefficient is lower than 0.6, the expert identification is still needed, and the results are reserved as new fault type judgment bases.
Further, in the first step, the normalization process specifically includes: zero-mean normalization is adopted, and the formula is as follows;
Figure BDA0002649011580000031
in the formula:
Figure BDA0002649011580000032
mean values for all data are indicated and σ indicates the standard deviation of all data.
Further, in the second step, after the cone crusher is started and stably operated, 30 groups of data are collected, and a group of approximate solutions are calculated by using a multiple linear regression method to serve as comparison data for fault diagnosis, specifically according to the following formula (2);
Figure BDA0002649011580000033
in the formula, xi,jI variable representing j group data, bnThe partial regression coefficient is expressed, n represents the group number, and the normal equation set can be further written into a matrix form, such as a formula;
Ab＝B (3)
wherein b is ═ b0,b1,…,bk-1]', introducing Y ═ xk1,xk2,…,xkn]', order
Figure BDA0002649011580000034
A＝XTX，
Figure BDA0002649011580000035
Solving equation (3) can obtain:
b＝A-1B＝(XTX)-1XTY (4)
and (3) acquiring 30 groups of stable operation data through a computer, and performing linear fitting according to a multiple linear regression theory to obtain a vector b for comparison and representing the normal state of the crusher.
Further, in the third step, a rolling updated partial regression coefficient vector b is calculated according to the multiple regression theorynewPerforming linear regression every 30 groups for subsequent sample data, each update group generating real-time updatesDetecting data;
the specific rolling update mode is as follows:
1, selecting 1 st to 30 th groups of measurement data to carry out 1 st operation;
selecting 2-31 groups of measurement data to perform 2 nd operation;
selecting the 3 rd to 32 th groups of measurement data to carry out the 3 rd operation;
fourthly, obtaining a plurality of groups of operation data by analogy.
Further, in the fourth step, for each newly generated detection data and the control data, a correlation coefficient is calculated, and the calculation formula of the correlation coefficient is as follows:
Figure BDA0002649011580000041
in the formula, bnewRespectively representing the vector representing the normal state of the cone crusher after linear regression operation and the vector representing the current state of the cone crusher after linear regression, Cov (b, b)new) Denotes b, bnewCovariance of (2), Var (b)new) Respectively represent b, bnewThe variance of (c).
Compared with the prior art, the invention has the beneficial effects that:
according to the fault diagnosis method research based on the linear regression crusher, the condition that the diagnosis result deviates from the field due to offline training is avoided, the expandability of data is improved by using a multivariate linear regression method, the interference possibly caused by instantaneous pulse signals is reduced by adopting a data processing mode of multiple groups of data and rolling updating, the fault diagnosis precision is continuously improved by continuously abundant fault judgment information, the production efficiency of enterprises is improved, and the burden of the enterprises is reduced.
Drawings
FIG. 1 is a block diagram of a fault diagnosis method for a crusher based on linear regression according to the present invention;
FIG. 2 is a schematic diagram of the data scrolling update according to the present invention.
Detailed Description
The following detailed description of the present invention will be made with reference to the accompanying drawings.
A fault diagnosis method based on a linear regression crusher specifically comprises the following steps:
1) selecting fault diagnosis parameters
As shown in figure 1, firstly determining input variables, selecting the main machine current of the cone crusher, the vibration amplitude of the transmission shaft, the vibration frequency of the transmission shaft, the vibration amplitude of the machine body, the vibration frequency of the machine body, the oil supply pressure of the lubricating oil pump, the tightening pressure, the release pressure and the lining plate pressure as initial fault diagnosis parameters, constructing a multi-element linear relation by taking the main machine current as a dependent variable and the other variables as independent variables, and enabling the amplitude of the transmission shaft to be x1The vibration frequency of the transmission shaft is x2Amplitude of fuselage x3The frequency of vibration of the fuselage being x4The oil supply pressure of the lubricating oil pump is x5With a locking pressure of x6Relief pressure x7Liner pressure is x8The host current is x9；
Selecting 179 groups of variable data, 1-81 groups of data as normal data, and 82-179 groups of data as abnormal data, and after the normalization by the formula (1), giving the first 30 groups of normalized measurement data as follows:
TABLE 1 first 30 sets of normalized variable data sheet
Figure BDA0002649011580000042
Figure BDA0002649011580000051
2) Generating control data for fault diagnosis
Calculating a partial regression coefficient vector representing the normal operation of the cone crusher by using a formula 2, wherein n is 30, k is 9, and the partial regression coefficient vector can be calculated by sequentially substituting data, specifically according to the following formula (2);
Figure BDA0002649011580000052
in the formula, xi,jI variable representing j group data, bnThe partial regression coefficients are expressed, n represents the number of groups, and the normal equation group can be further written into a matrix form, such as formula (3);
Ab＝B (3)
wherein b is ═ b0,b1,…,bk-1]', introducing Y ═ xk1,xk2,…,xkn]', order
Figure BDA0002649011580000061
A＝XTX，
Figure BDA0002649011580000062
The solution of the formula (3) can be obtained,
b＝A-1B＝(XTX)-1XTY (4)
obtaining 30 groups of stable operation data through a computer and carrying out linear fitting according to a multiple linear regression theory to obtain a vector for comparison and representing normal state of the crusher
b＝[-0.6463,0.2497,-0.1031,-0.0399,0.0510,0.0400,0.1431,0.0040]′。
3) Rolling to generate detection data representing current state of crusher
As shown in FIG. 2, after the cone crusher generates the control data for representing the normal state, linear regression is carried out once for every 30 groups of subsequent sampling data, each time one group is updated to generate real-time updated detection data, and the 31 st group of data is calculated
bnew＝[-0.8352,0.1122,0.0170,-0.0341,-0.0501,0.0852,0.0653,-0.0338]，
b,bnewRespectively representing the characterization cones after linear regression operationThe vector of the normal state of the crusher and the vector after linear regression representing the current state of the cone crusher are updated in real time, wherein the 31 th group of data refers to data used in the 31 st operation, and the following steps are the same;
4) calculating the correlation of the test data and the control data
For each newly generated test data, a correlation coefficient calculation is performed with the control data, the correlation coefficient is calculated as follows,
Figure BDA0002649011580000063
in the formula, bnewRespectively representing the vector representing the normal state of the cone crusher after linear regression operation and the vector representing the current state of the cone crusher after linear regression, Cov (b, b)new) Denotes b, bnewCovariance of (2), Var (b)new) Respectively represent b, bnewThe variance of (a);
partial regression coefficient vector b calculated from group 31 datanewThe calculation result of the correlation coefficient of the partial regression coefficient vector b calculated with the group 1 data is r which is 0.9485, and the current running state of the crusher is stable;
5) obtain the diagnosis result and enrich the basis of the next diagnosis
For the crusher diagnosed with the abnormal state, after expert identification, fault types are recorded, a group of multiple regression partial regression coefficient vectors generated by real-time data when the correlation coefficient is lower than 0.6 is reserved, the partial regression coefficient vectors are used for judging fault types when faults occur next time, if the faults are not of the same type, namely the correlation coefficient of the partial regression coefficient vectors with the known faults is lower than 0.6, the expert identification is still required, and the results are reserved as a new fault type judgment basis;
if the correlation coefficients of the group 66 data to the group 75 data are 0.5643, 0.6037, 0.6078, 0.6067, 0.5989, 0.5670, 0.4490, 0.4683, 0.4471, 0.4803, 0.5441, it indicates that the deviation from the normal operation range is detected after the shutdown.
The above embodiments are implemented on the premise of the technical solution of the present invention, and detailed embodiments and specific operation procedures are given, but the scope of the present invention is not limited to the above embodiments. The methods used in the above examples are conventional methods unless otherwise specified.

Claims (5)

1. A fault diagnosis method based on a linear regression crusher is characterized in that host current, transmission shaft amplitude and frequency, and body amplitude and frequency of a cone crusher are selected as fault diagnosis parameters, n groups of data recorded after the cone crusher works stably are established into a partial regression coefficient vector for diagnosis by using a multiple regression method after normalization processing, new data are subjected to one-time multiple linear regression by using the n groups to obtain a vector representing the current state, and a correlation coefficient is calculated between the current vector and the diagnosis vector so as to reflect the degree of the current operation state of the cone crusher deviating from the stable state in normal operation; when the correlation degree is below 0.6 but no sharp descending trend appears, the machine is checked after shutdown, the fault type is recorded according to the check result, and the array after multivariate regression is added into a comparison database for next fault diagnosis; when the correlation degree is lower than 0.6 and a rapid descending trend continues to appear, the machine should be stopped for checking as soon as possible; the method specifically comprises the following steps:
step one, selecting fault diagnosis parameters and carrying out normalization processing
According to the working characteristics of the cone crusher, the current signal and the vibration signal are rapidly changed when the crusher stably operates, and the cone crusher has the capability of rapidly reflecting the current operation condition of equipment relative to signals such as temperature, oil pressure and the like, so that the host machine current, the transmission shaft amplitude and frequency, the machine body amplitude and frequency, the lubricating oil tank oil supply pressure, the locking pressure, the releasing pressure and the lining plate pressure of the cone crusher are selected as the basis of fault diagnosis;
step two, generating comparison data for fault diagnosis
Acquiring n groups of data after the cone crusher is started and stably operates, and calculating a group of approximate solutions by using a multiple linear regression method to serve as comparison data for fault diagnosis; obtaining n groups of stable operation data through a computer, and performing linear fitting according to a multiple linear regression theory to obtain a vector b for comparison and representing normal state of the crusher;
step three, generating detection data representing the current state of the crusher in a rolling manner
After the cone crusher generates the contrast data for representing the normal state, the rolling updated partial regression coefficient vector b is calculated according to the multiple regression theorynewPerforming linear regression on each n groups of subsequent sampling data, and updating one group of the subsequent sampling data each time to generate real-time updated detection data;
step four, calculating the correlation between the detection data and the comparison data
B, b is calculated according to a correlation coefficient calculation formulanewThe correlation coefficient r between the two is judged according to the threshold value of 0.6 to judge whether the current running state of the crusher is normal, r>0.6 indicates that the crusher is normally operated, and when the correlation coefficient is reduced to be below 0.6, the cone crusher is in a state of deviating from the normal operation seriously, and should be overhauled after shutdown;
step five, obtaining a diagnosis result and taking the diagnosis result as the basis of the next diagnosis
For the crusher diagnosed with the abnormal state, after expert identification, fault types are recorded, a group of partial regression coefficient vectors after multiple regression generated by real-time data when the correlation coefficient is lower than 0.6 is reserved, the partial regression coefficient vectors are used for judging fault types when faults occur next time, if the partial regression coefficient vectors are not the same type of faults, namely the correlation coefficient is lower than 0.6, the expert identification is still needed, and the results are reserved as new fault type judgment bases.
2. The method for diagnosing the fault of the crusher based on the linear regression as claimed in claim 1, wherein in the first step, the normalization process is specifically as follows: zero-mean normalization is adopted, and the formula is as follows;
Figure FDA0002649011570000021
in the formula:
Figure FDA0002649011570000022
mean values for all data are indicated and σ indicates the standard deviation of all data.
3. The fault diagnosis method based on the linear regression crusher as claimed in claim 1, wherein in the second step, 30 groups of data are collected after the cone crusher is started and stably operated, and a group of approximate solutions are calculated by using a multiple linear regression method as comparison data for fault diagnosis, specifically according to the following formula (2);
Figure FDA0002649011570000023
in the formula, xi,jI variable representing j group data, bnThe partial regression coefficient is expressed, n represents the group number, and the normal equation set can be further written into a matrix form, such as a formula;
Ab＝B (3)
wherein b is ═ b0,b1,…,bk-1]', introducing Y ═ xk1,xk2,…,xkn]', order
Figure FDA0002649011570000024
A＝XTX，
Figure FDA0002649011570000025
Solving equation (3) can obtain:
b＝A-1B＝(XTX)-1XTY (4)
and taking n as 30, obtaining 30 groups of stable operation data through a computer, and performing linear fitting according to a multiple linear regression theory to obtain a vector b for comparison and representing the normal state of the crusher.
4. The method for fault diagnosis of crusher based on linear regression as claimed in claim 1, wherein in step three, the rolling updated partial regression coefficient vector b is calculated according to multiple regression theorynewPerforming linear regression on each 30 groups of subsequent sampling data, and updating one group of the subsequent sampling data each time to generate real-time updated detection data;
the specific rolling update mode is as follows:
1, selecting 1 st to 30 th groups of measurement data to carry out 1 st operation;
selecting 2-31 groups of measurement data to perform 2 nd operation;
selecting the 3 rd to 32 th groups of measurement data to carry out the 3 rd operation;
fourthly, obtaining a plurality of groups of operation data by analogy.
5. The method for diagnosing the faults of the crusher based on the linear regression as claimed in claim 1, wherein in the fourth step, the correlation coefficient calculation is performed on each newly generated detection data and the control data, and the calculation formula of the correlation coefficient is as follows:
Figure FDA0002649011570000031
in the formula, bnewRespectively representing the vector representing the normal state of the cone crusher after linear regression operation and the vector representing the current state of the cone crusher after linear regression, Cov (b, b)new) Denotes b, bnewCovariance of (2), Var (b)new) Respectively represent b, bnewThe variance of (c).
CN202010863644.9A 2020-08-25 2020-08-25 Crusher fault diagnosis method based on linear regression Active CN112090478B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010863644.9A CN112090478B (en) 2020-08-25 2020-08-25 Crusher fault diagnosis method based on linear regression

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010863644.9A CN112090478B (en) 2020-08-25 2020-08-25 Crusher fault diagnosis method based on linear regression

Publications (2)

Publication Number Publication Date
CN112090478A true CN112090478A (en) 2020-12-18
CN112090478B CN112090478B (en) 2022-06-17

Family

ID=73753290

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010863644.9A Active CN112090478B (en) 2020-08-25 2020-08-25 Crusher fault diagnosis method based on linear regression

Country Status (1)

Country Link
CN (1) CN112090478B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114308358A (en) * 2022-03-17 2022-04-12 山东金有粮脱皮制粉设备有限公司 Safe operation monitoring system of corncob grinding device
CN114398800A (en) * 2022-01-27 2022-04-26 矿冶科技集团有限公司 Fault diagnosis method and device for crusher system

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030040878A1 (en) * 2001-08-09 2003-02-27 Rovsing Dynamics A/S Automatic machinery fault diagnostic method and apparatus
CN101269350A (en) * 2008-04-25 2008-09-24 东南大学 Operation optimizing and energy-saving control method for intermediate storage iron shot mill flour milling system
CN201244509Y (en) * 2008-05-27 2009-05-27 广东远大机械集团有限公司 Intelligent device for monitoring state of gear drive of spindle breaker
CN105221307A (en) * 2014-06-27 2016-01-06 Fpt发动机研究公司 For the system of the detection air filter situation of internal-combustion engine
CN204953093U (en) * 2015-08-30 2016-01-13 宋国丰 High efficiency cone crusher monitoring system
CN106202635A (en) * 2016-06-28 2016-12-07 西安理工大学 A kind of dynamic axle temperature Forecasting Methodology of bullet train based on multivariate regression models
CN107127043A (en) * 2017-06-28 2017-09-05 中国科学院上海高等研究院 A kind of disintegrating machine fault monitoring system and method
WO2019098438A1 (en) * 2017-11-16 2019-05-23 삼영플랜트 주식회사 Lubricating system of intelligent cone crusher
CN110967182A (en) * 2019-11-13 2020-04-07 鞍钢集团矿业有限公司 Cone crusher vibration data acquisition and preprocessing method
CN211190340U (en) * 2019-06-17 2020-08-07 太原重工股份有限公司 Double-geared roller crusher state monitoring and fault diagnosis system

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030040878A1 (en) * 2001-08-09 2003-02-27 Rovsing Dynamics A/S Automatic machinery fault diagnostic method and apparatus
CN101269350A (en) * 2008-04-25 2008-09-24 东南大学 Operation optimizing and energy-saving control method for intermediate storage iron shot mill flour milling system
CN201244509Y (en) * 2008-05-27 2009-05-27 广东远大机械集团有限公司 Intelligent device for monitoring state of gear drive of spindle breaker
CN105221307A (en) * 2014-06-27 2016-01-06 Fpt发动机研究公司 For the system of the detection air filter situation of internal-combustion engine
CN204953093U (en) * 2015-08-30 2016-01-13 宋国丰 High efficiency cone crusher monitoring system
CN106202635A (en) * 2016-06-28 2016-12-07 西安理工大学 A kind of dynamic axle temperature Forecasting Methodology of bullet train based on multivariate regression models
CN107127043A (en) * 2017-06-28 2017-09-05 中国科学院上海高等研究院 A kind of disintegrating machine fault monitoring system and method
WO2019098438A1 (en) * 2017-11-16 2019-05-23 삼영플랜트 주식회사 Lubricating system of intelligent cone crusher
CN211190340U (en) * 2019-06-17 2020-08-07 太原重工股份有限公司 Double-geared roller crusher state monitoring and fault diagnosis system
CN110967182A (en) * 2019-11-13 2020-04-07 鞍钢集团矿业有限公司 Cone crusher vibration data acquisition and preprocessing method

Non-Patent Citations (6)

* Cited by examiner, † Cited by third party
Title
周光海等: "基于核偏最小二乘法的物联网无线传感网络故障分析与研究", 《通信学报》 *
水鑫忠: "矿用破碎机状态监测系统研究", 《内蒙古煤炭经济》 *
贺湘宇等: "基于偏最小二乘回归的挖掘机液压系统故障诊断", 《中南大学学报(自然科学版)》 *
陈敏维等: "偏最小二乘回归法在介损在线监测误差分析中的应用", 《电力系统及其自动化学报》 *
顾清华等: "金属露天矿旋回式破碎机在线故障诊断模型构建与实时分析", 《金属矿山》 *
高田静: "浅谈旋回破碎机故障诊断与保养", 《内燃机与配件》 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114398800A (en) * 2022-01-27 2022-04-26 矿冶科技集团有限公司 Fault diagnosis method and device for crusher system
CN114398800B (en) * 2022-01-27 2022-10-25 矿冶科技集团有限公司 Fault diagnosis method and device for crusher system
CN114308358A (en) * 2022-03-17 2022-04-12 山东金有粮脱皮制粉设备有限公司 Safe operation monitoring system of corncob grinding device
CN114308358B (en) * 2022-03-17 2022-05-27 山东金有粮脱皮制粉设备有限公司 Safe operation monitoring system of corncob grinding device

Also Published As

Publication number Publication date
CN112090478B (en) 2022-06-17

Similar Documents

Publication Publication Date Title
CN112090478B (en) Crusher fault diagnosis method based on linear regression
CN109766334B (en) Processing method and system for online monitoring abnormal data of power equipment
US7536371B2 (en) Apparatus and method for the analysis of a process having parameter-based faults
US8010321B2 (en) Metrics independent and recipe independent fault classes
US9547544B2 (en) Method for verifying bad pattern in time series sensing data and apparatus thereof
CN112036089A (en) Coal mill fault early warning method based on DPC-MND and multivariate state estimation
CN115425764A (en) Real-time monitoring method, system and storage medium for intelligent network risk of electric power system
US20170307480A1 (en) Data Display System
CN104360677A (en) Cigarette processing quality evaluation and diagnosis method
CN111752147B (en) Multi-working-condition process monitoring method with continuous learning capability and improved PCA (principal component analysis)
CN111352408B (en) Multi-working-condition process industrial process fault detection method based on evidence K nearest neighbor
Li et al. Canonical variate residuals-based contribution map for slowly evolving faults
CN107153748A (en) Based on weighting core pivot element analysis（WKPCA）Rotary kiln method for diagnosing faults
CN109523030B (en) Telemetering parameter abnormity monitoring system based on machine learning
CN110083797A (en) A kind of drip irrigation pipe production line exception method of real-time and system
CN105759787B (en) A kind of method for diagnosing faults having supervision linear dynamic system model based on switching
JPH04190133A (en) Diagnosis support apparatus of equipment
CN111914888A (en) Chemical process monitoring method integrating multi-working-condition identification and fault detection
CN107272625B (en) A kind of industrial process method for diagnosing faults based on bayesian theory
CN111861181A (en) Intelligent production system and production line formulation method for digital energy products
CN116961215A (en) Rapid fault response processing method for power system
CN110347138A (en) The Fault Diagnosis for Chemical Process method of self-adaptive kernel pivot analysis
CN111624985B (en) Gas turbine control system sensor fault diagnosis method
CN115327436A (en) Power distribution network short-circuit fault diagnosis method based on composite statistics
Li et al. A LS-SVM based approach for turbine engines prognostics using sensor data

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant