CN110108457B - Primary fan shaft temperature diagnosis method based on universal gravitation neural network - Google Patents

Primary fan shaft temperature diagnosis method based on universal gravitation neural network Download PDF

Info

Publication number
CN110108457B
CN110108457B CN201910313615.2A CN201910313615A CN110108457B CN 110108457 B CN110108457 B CN 110108457B CN 201910313615 A CN201910313615 A CN 201910313615A CN 110108457 B CN110108457 B CN 110108457B
Authority
CN
China
Prior art keywords
primary air
air fan
primary
neural network
fan
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.)
Active
Application number
CN201910313615.2A
Other languages
Chinese (zh)
Other versions
CN110108457A (en
Inventor
马博洋
李金拓
屈广顺
马文元
李强
董蔚
王文铁
齐兵
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.)
Harbin No1 Thermoelectric Plant Of Datang Heilongjiang Power Generation Co Ltd
Original Assignee
Harbin No1 Thermal Power Plant Datang Heilongjiang Power Generation Co ltd
Datang Northeast Electric Power Test and Research Institute Co Ltd
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 Harbin No1 Thermal Power Plant Datang Heilongjiang Power Generation Co ltd, Datang Northeast Electric Power Test and Research Institute Co Ltd filed Critical Harbin No1 Thermal Power Plant Datang Heilongjiang Power Generation Co ltd
Priority to CN201910313615.2A priority Critical patent/CN110108457B/en
Publication of CN110108457A publication Critical patent/CN110108457A/en
Application granted granted Critical
Publication of CN110108457B publication Critical patent/CN110108457B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M13/00—Testing of machine parts
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M13/00—Testing of machine parts
    • G01M13/04—Bearings
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Molecular Biology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Testing And Monitoring For Control Systems (AREA)

Abstract

The invention relates to a primary fan shaft temperature diagnosis method based on a universal gravitation neural network, which comprises the following steps: step 1, establishing and training a corresponding BP neural network prediction model based on relevant measuring point parameters of a primary air fan, and optimizing the BP neural network prediction model based on a universal gravitation search algorithm to obtain the universal gravitation neural network prediction model; step 2, estimating the shaft temperature of the primary air fan based on the universal gravitation neural network prediction model to obtain the predicted values of the front, middle and rear shaft temperatures of the primary air fan; and 3, solving a residual error between the real shaft temperature value and the predicted shaft temperature value based on the MSET algorithm, and diagnosing a fault if the change trend of the residual error curve is continuous and has no fall-back trend. The early warning method can realize early fault early warning of the primary fan shaft temperature and has higher accuracy.

Description

Primary fan shaft temperature diagnosis method based on universal gravitation neural network
Technical Field
The invention belongs to the technical field of thermal power generation, and particularly relates to a primary fan shaft temperature diagnosis method based on a universal gravitation neural network.
Background
With the continuous development of thermal power generation high-parameter and large-capacity units, the auxiliary machines corresponding to the thermal power generation high-parameter and large-capacity units tend to be large-sized, and the functions of auxiliary equipment in the production process of thermal power plants are more and more important. The failure of the auxiliary equipment not only directly affects the output and quality of electric energy, but also can cause serious equipment and personal accidents. Therefore, the method has important practical significance for monitoring the state and diagnosing the fault of the auxiliary equipment of the thermal power plant.
Currently, there are three types of commonly used online monitoring and fault diagnosis methods based on auxiliary equipment of a thermal power plant:
(1) and (4) analyzing the model method.
The state monitoring method based on the analytical model is time-consuming and labor-consuming, is limited by accurate modeling due to the problems of time variation, uncertainty and the like of the system, and is not suitable for a system with a complex structure and strong nonlinearity, such as auxiliary equipment of a thermal power plant.
(2) An expert system method.
The expert system designs a group of computer programs by using expert experience, and carries out intelligent analysis based on the expert experience on the monitored state, thereby carrying out fault judgment. The method has wide application to diagnosis of nonlinear systems and complex systems, but researches show that the method of the expert system has the following main problems in fault diagnosis of auxiliary equipment of the thermal power plant: in practical applications, a "bottleneck" exists in knowledge acquisition; the dynamic process information of the running process is not fully utilized; new faults and miscellaneous faults unknown to the system are not adequately handled.
(3) Machine learning and pattern recognition methods.
The core of the fault diagnosis based on machine learning is to reasonably process and acquire fault information and perform accurate pattern recognition and state prediction on a diagnosis object under a specific environment. Due to the restriction of the current machine learning bottleneck problem, the diagnosis object with strong randomness, strong uncertainty and redundant or deficient knowledge information is still difficult to have higher diagnosis accuracy.
Disclosure of Invention
The invention aims to provide a primary fan shaft temperature diagnosis method based on a universal gravitation neural network.
The invention provides a primary fan shaft temperature diagnosis method based on a universal gravitation neural network, which comprises the following steps:
step 1, establishing and training a corresponding BP neural network prediction model based on relevant measuring point parameters of a primary air fan, and optimizing the BP neural network prediction model based on a universal gravitation search algorithm to obtain the universal gravitation neural network prediction model;
step 2, estimating the shaft temperature of the primary air fan based on the universal gravitation neural network prediction model to obtain the predicted values of the front, middle and rear shaft temperatures of the primary air fan;
and 3, solving a residual error between the real shaft temperature value and the predicted shaft temperature value based on the MSET algorithm, and diagnosing a fault if the change trend of the residual error curve is continuous and has no fall-back trend.
Further, the primary air fan related measuring point parameters in the step 1 include multiple kinds of primary air fan oil pump current, primary air fan outlet pressure, primary air fan vertical vibration, primary air fan motor current, primary air fan motor stator temperature, primary air fan motor rear bearing temperature, primary air fan motor front bearing temperature, primary air fan movable blade control feedback, primary air fan air quantity, primary air fan control oil pressure, primary air fan inlet pressure, primary air fan stall alarm differential pressure, primary air fan horizontal vibration, primary air fan oil station oil tank temperature and primary air fan middle bearing temperature.
Further, primary fan oil pump current, primary fan outlet pressure, primary fan vertical vibration, primary fan motor current, primary fan motor stator temperature, primary fan motor rear bearing temperature, primary fan motor front bearing temperature, primary fan movable blade control feedback, primary fan air volume, primary fan rear bearing temperature, primary fan control oil pressure, primary fan front bearing temperature, primary fan inlet pressure, primary fan stall alarm differential pressure, primary fan horizontal vibration, primary fan oil station oil tank temperature are input parameters of the universal gravitation neural network prediction model; the temperature of the front bearing of the primary air fan, the temperature of the middle bearing of the primary air fan and the temperature of the rear bearing of the primary air fan are output parameters of the universal gravitation neural network prediction model.
By means of the scheme, the primary fan shaft temperature diagnosis method based on the universal gravitation neural network has the following technical effects:
1) the method comprises the steps of estimating the shaft temperature of the primary air fan by utilizing a plurality of measuring point data related to the primary air fan acquired by a power plant SIS system, carrying out MSET (minimum shift event) on the estimated shaft temperature and the actual shaft temperature of the primary air fan to obtain a residual error, judging the trend of the residual error and whether the residual error is suddenly changed, updating and early warning in real time, facilitating a maintainer to know the state of equipment in advance, and carrying out detailed inspection according to warning;
2) parameters of the neural network prediction model are optimized by utilizing a universal gravitation algorithm, so that the estimation precision of the prediction model can be effectively improved;
3) the MSET algorithm is used for calculating residual errors, so that the accuracy of the model is improved;
4) by selecting more variables, the coverage degree of the model is improved.
The foregoing description is only an overview of the technical solutions of the present invention, and in order to make the technical solutions of the present invention more clearly understood and to implement them in accordance with the contents of the description, the following detailed description is given with reference to the preferred embodiments of the present invention and the accompanying drawings.
Drawings
FIG. 1 is a flow chart of a primary fan shaft temperature diagnosis method based on a universal gravitation neural network;
FIG. 2 is a diagram of a model architecture of the universal gravitation neural network of the present invention;
FIG. 3 is a three-layer BP neural network structure diagram adopted by the universal gravitation neural network model of the present invention;
FIG. 4 is a flowchart of optimizing a BP neural network prediction model using a universal gravitation search algorithm according to the present invention.
Detailed Description
The following detailed description of embodiments of the present invention is provided in connection with the accompanying drawings and examples. The following examples are intended to illustrate the invention but are not intended to limit the scope of the invention.
Referring to fig. 1, the embodiment provides a primary air fan shaft temperature diagnosis method based on a universal gravitation neural network, including the following steps:
s1, establishing and training a corresponding BP neural network prediction model based on the primary fan related measuring point parameters, and optimizing the BP neural network prediction model based on the universal gravitation search algorithm to obtain the universal gravitation neural network prediction model;
s2, estimating the primary fan shaft temperature based on the universal gravitation neural network prediction model to obtain the predicted values of the front, middle and rear shaft temperatures of the primary fan;
and step S3, obtaining a residual error between the real shaft temperature value and the predicted shaft temperature value based on the MSET algorithm, and diagnosing a fault if the change trend of the residual error curve is continuous and has no fall-back trend.
By the primary fan shaft temperature diagnosis method based on the universal gravitation neural network, early fault early warning of the primary fan shaft temperature can be achieved, and high accuracy is achieved.
In this embodiment, the primary air fan related measurement point parameters in step S1 include a plurality of primary air fan oil pump current, primary air fan outlet pressure, primary air fan vertical vibration, primary air fan motor current, primary air fan motor stator temperature, primary air fan motor rear bearing temperature, primary air fan motor front bearing temperature, primary air fan movable blade control feedback, primary air fan air volume, primary air fan rear bearing temperature, primary air fan control oil pressure, primary air fan front bearing temperature, primary air fan inlet pressure, primary air fan stall alarm differential pressure, primary air fan horizontal vibration, primary air fan oil station oil tank temperature, and primary air fan middle bearing temperature.
In this embodiment, a primary fan oil pump current, a primary fan outlet pressure, a primary fan vertical vibration, a primary fan motor current, a primary fan motor stator temperature, a primary fan motor rear bearing temperature, a primary fan motor front bearing temperature, a primary fan movable blade control feedback, a primary fan air volume, a primary fan control oil pressure, a primary fan inlet pressure, a primary fan stall alarm differential pressure, a primary fan horizontal vibration, a primary fan oil station oil tank temperature are input parameters of the universal gravitation neural network prediction model; the temperature of the primary air fan front bearing, the temperature of the primary air fan middle bearing and the temperature of the primary air fan rear bearing are output parameters of the universal gravitation neural network prediction model.
The present invention is described in further detail below.
Referring to fig. 2 to 4, in an embodiment, a method for performing primary fan shaft temperature diagnosis by using the method includes:
step 1: according to the related measuring points of the primary fan, the number of the measuring points is 35, and the measuring points are respectively the primary fan #1 oil pump current; primary blower #2 oil pump current; primary fan outlet pressure; the primary fan vertically vibrates 1; the primary fan vibrates vertically 2; primary fan motor current; the temperature of a stator A of a primary fan motor is 1; the temperature of a stator A of a primary fan motor is 2; the temperature of a stator B of a primary fan motor is 1; the temperature of a stator B of a primary fan motor is 2; the temperature of a primary fan motor stator C is 1; the temperature of a primary fan motor stator C is 2; the temperature of a rear bearing of a primary fan motor; the temperature of a front bearing of a primary fan motor; controlling a movable blade of a primary fan; the movable blade of the primary fan controls and feeds back; the movable blades of the primary fan control output; offset of a movable blade of the primary fan; feeding back the position of a movable blade of the primary fan; the air quantity of a primary fan; the temperature of a rear bearing of the primary air fan is 1; the temperature of a rear bearing of the primary air fan is 2; the temperature of the rear bearing of the primary air fan is 3; the primary fan controls the oil pressure to be normal; the temperature of a front bearing of the primary air fan is 1; the temperature of a front bearing of the primary air fan is 2; the temperature of a front bearing of the primary air fan is 3; primary fan inlet pressure; a primary fan stalls and alarms differential pressure; horizontally vibrating a primary fan by 1; horizontally vibrating a primary fan 2; the temperature of an oil tank of a primary air fan oil station and the temperature of a bearing in the primary air fan are 1; the bearing temperature in the primary air fan is 2; establishing and training a corresponding universal gravitation neural network prediction model for the temperature 3 of a bearing in the primary air fan, wherein the output of the model is the temperature of a front shaft, a middle shaft and a rear shaft of the primary air fan, and the input of the model is other variables, and the method comprises the following steps:
a. collecting data to construct a sample set;
recording the running record of a primary fan at a certain moment k, and recording the variable values of all 35 related measuring points to obtain a sample (C)0(k)，C1(k)，…,C31(k) ); forming a sample set by collecting samples at a plurality of different time instances (C)0(k)，C1(k)，…,C31(k))}；
b. Establishing a BP neural network prediction model with a three-layer structure, wherein the number of input layer neuron nodes is 32, the number of hidden layer neuron nodes is 42, the number of output layer neuron nodes is 3, a hyperbolic tangent function is used as a hidden layer neuron transfer function, and an S-type function is used as an output layer neuron transfer function; for the sample (C) constructed at the k-th time instant0(k)，C1(k)，…,C31(k) Taking the first output of the BP neural network prediction model as T1(k) The second output is T2(k) And the third output is T3(k)；
c. Randomly extracting 80% of a sample set as a training sample, taking the rest 20% as an inspection sample, and optimizing the weight and the threshold of the established BP neural network prediction model by using a universal gravitation search algorithm, wherein the specific steps are as follows:
setting a particle group size N and an initial position of each particle
Figure BDA0002032327120000051
Figure BDA0002032327120000052
D is a particle dimension, and the position of each particle is initialized by adopting a random number generation mode;
calculating a fitness function value of each particle:
defining a fitness function f of a particleiMean square error of the model for the BP neural network prediction on the training samples:
Figure BDA0002032327120000053
wherein: m is the number of output nodes; p is the number of training samples;
Figure BDA0002032327120000061
an expected output value for the network; x is the number ofpjActual output values for the network;
③ updating f in the groupbestAnd fworst(fbest＝minfj，fworst＝maxfj) The mass M of each particle was calculated as followsi：
Figure BDA0002032327120000062
Figure BDA0002032327120000063
Fourthly, calculating the resultant force F of the gravitation of each particle according to the following formulaiAnd acceleration ai：
Figure BDA0002032327120000064
Figure BDA0002032327120000065
Figure BDA0002032327120000066
In the formula: t is the number of iterations, Fi d(t) the resultant force F of the attraction from other particles to the ith particleiThe component of the d-th dimension of (c),
Figure BDA0002032327120000067
acceleration of the ith particle in d dimension, randjIs [0, 1 ]]Random number between, G (t) is the gravitational time constantCounting; mpi(t) and Maj(t) the passive inertial mass of the ith particle and the active inertial mass of the jth particle, respectively,
Figure BDA0002032327120000068
and
Figure BDA0002032327120000069
the d-dimension positions of the ith particle and the jth particle respectively;
updating the velocity v of each particle according to the following formulaiAnd position Pi：
Figure BDA00020323271200000610
Figure BDA00020323271200000611
In the formula:
Figure BDA0002032327120000071
d-dimension velocity of the ith particle;
sixthly, returning to the step II, stopping iteration after reaching the maximum iteration times, and at the moment fbestThe corresponding particle positions are the weight and the threshold of the optimized BP neural network prediction model;
step 2: and (3) estimating the shaft temperature of the primary air fan by using the optimized universal gravitation neural network prediction model, namely, taking 32 variables related to the primary air fan as the input of the model, wherein the output of the model is the front, middle and rear shaft temperatures of the primary air fan.
And step 3: and (3) solving a residual error between the real shaft temperature value and the predicted shaft temperature value by using an MSET (modeling and optimization) model, paying attention to the change trend of a residual error curve, if the change trend is continuously increased, informing maintenance personnel to confirm the running state of the primary air fan, and if a fault exists, taking corresponding measures. The algorithm operation speed can be controlled to be refreshed once in 15 seconds, and the real-time fault early warning is basically met.
The invention has the following technical effects:
1) the method comprises the steps of estimating the shaft temperature of the primary air fan by utilizing a plurality of measuring point data related to the primary air fan acquired by a power plant SIS system, carrying out MSET (minimum shift event) on the estimated shaft temperature and the actual shaft temperature of the primary air fan to obtain a residual error, judging the trend of the residual error and whether the residual error is suddenly changed, updating and early warning in real time, facilitating a maintainer to know the state of equipment in advance, and carrying out detailed inspection according to warning;
2) parameters of the neural network prediction model are optimized by utilizing a universal gravitation algorithm, so that the estimation precision of the prediction model can be effectively improved;
3) the MSET algorithm is used for calculating residual errors, so that the accuracy of the model is improved;
4) by selecting more variables, the coverage degree of the model is improved.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, it should be noted that, for those skilled in the art, many modifications and variations can be made without departing from the technical principle of the present invention, and these modifications and variations should also be regarded as the protection scope of the present invention.

Claims (3)

1. A primary air fan shaft temperature diagnosis method based on a universal gravitation neural network is characterized by comprising the following steps:
step 1, establishing a universal gravitation neural network prediction model based on data of relevant measuring points of a primary air fan, and optimizing the universal gravitation neural network prediction model;
the data of the primary air fan related measuring points comprise:
primary air fan oil pump current data, primary air fan outlet pressure data, primary air fan vertical vibration data, primary air fan motor current data, primary air fan motor stator temperature data, primary air fan motor rear bearing temperature data, primary air fan motor front bearing temperature data, primary air fan movable blade control feedback data, primary air fan movable blade control output data, primary air fan movable blade offset data, primary air fan movable blade position feedback data, primary air fan air volume data, primary air fan rear bearing temperature data, primary air fan control oil pressure normal data, primary air fan inlet pressure data, primary air fan stall alarm differential pressure data, primary air fan horizontal vibration data, primary air fan oil station oil tank temperature data and primary air fan middle bearing temperature data;
the method for establishing and optimizing the universal gravitation neural network prediction model comprises the following steps:
a. collecting data to construct a sample set;
recording the running record of a primary fan at a certain moment k, and recording the variable values of all 35 related measuring points to obtain a sample (C)0(k)，C1(k)，…,C31(k) ); forming a sample set by collecting samples at a plurality of different time instances (C)0(k)，C1(k)，…,C31(k))}；
b. Establishing a BP neural network prediction model with a three-layer structure, wherein the number of input layer neuron nodes is 32, the number of hidden layer neuron nodes is 42, the number of output layer neuron nodes is 3, a hyperbolic tangent function is used as a hidden layer neuron transfer function, and an S-type function is used as an output layer neuron transfer function; for the sample (C) constructed at the k-th time instant0(k)，C1(k)，…,C31(k) Taking the first output of the BP neural network prediction model as T1(k) The second output is T2(k) And the third output is T3(k)；
c. Randomly extracting 80% of a sample set as a training sample, taking the rest 20% as an inspection sample, and optimizing the weight and the threshold of the established BP neural network prediction model by using a universal gravitation search algorithm, wherein the specific steps are as follows:
setting a particle group size N and an initial position of each particle
Figure FDA0003230826060000011
Figure FDA0003230826060000012
D is a particle dimension, and the position of each particle is initialized by adopting a random number generation mode;
calculating a fitness function value of each particle:
defining a fitness function f of a particleiMean square error of the model for the BP neural network prediction on the training samples:
Figure FDA0003230826060000021
wherein: m is the number of output nodes; p is the number of training samples;
Figure FDA0003230826060000022
an expected output value for the network; x is the number ofpjActual output values for the network;
③ updating f in the groupbestAnd fworst(fbest＝minfj，fworst＝maxfj) The mass M of each particle was calculated as followsi：
Figure FDA0003230826060000023
Figure FDA0003230826060000024
Fourthly, calculating the resultant force F of the gravitation of each particle according to the following formulaiAnd acceleration ai：
Figure FDA0003230826060000025
Figure FDA0003230826060000026
Figure FDA0003230826060000027
In the formula: t is the number of iterations, Fi d(t) isiThe resultant force F of attraction of the particles from other particlesiThe component of the d-th dimension of (c),
Figure FDA0003230826060000028
acceleration of the ith particle in d dimension, randjIs [0, 1 ]]G (t) is the gravitational time constant; mpi(t) and Maj(t) the passive inertial mass of the ith particle and the active inertial mass of the jth particle, respectively,
Figure FDA0003230826060000031
and
Figure FDA0003230826060000032
the d-dimension positions of the ith particle and the jth particle respectively;
updating the velocity v of each particle according to the following formulaiAnd position Pi：
Figure FDA0003230826060000033
Figure FDA0003230826060000034
In the formula:
Figure FDA0003230826060000035
d-dimension velocity of the ith particle;
sixthly, returning to the step II, stopping iteration after reaching the maximum iteration times, and at the moment fbestThe corresponding particle positions are the weight and the threshold of the optimized BP neural network prediction model;
step 2, estimating the shaft temperature of the primary fan by using the optimized universal gravitation neural network prediction model to obtain a predicted shaft temperature value;
and 3, solving a residual error between the real shaft temperature value and the predicted shaft temperature value based on the MSET algorithm, diagnosing the shaft temperature of the primary fan according to the change trend of a residual error curve, and judging that the shaft temperature of the primary fan is abnormal if the change trend of the residual error curve is continuously increased.
2. The primary fan shaft temperature diagnosis method based on the universal gravitation neural network as claimed in claim 1, further comprising:
and carrying out real-time early warning on the bearing fault of the primary fan based on the shaft temperature diagnosis result of the primary fan.
3. The primary wind turbine shaft temperature diagnosis method based on the universal gravitation neural network as claimed in claim 2, wherein in the step 3, the operation speed of the MSET algorithm is controlled to be refreshed once in 15 seconds.
CN201910313615.2A 2019-04-18 2019-04-18 Primary fan shaft temperature diagnosis method based on universal gravitation neural network Active CN110108457B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910313615.2A CN110108457B (en) 2019-04-18 2019-04-18 Primary fan shaft temperature diagnosis method based on universal gravitation neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910313615.2A CN110108457B (en) 2019-04-18 2019-04-18 Primary fan shaft temperature diagnosis method based on universal gravitation neural network

Publications (2)

Publication Number Publication Date
CN110108457A CN110108457A (en) 2019-08-09
CN110108457B true CN110108457B (en) 2021-11-09

Family

ID=67485813

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910313615.2A Active CN110108457B (en) 2019-04-18 2019-04-18 Primary fan shaft temperature diagnosis method based on universal gravitation neural network

Country Status (1)

Country Link
CN (1) CN110108457B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112414694B (en) * 2020-06-12 2021-08-27 北京航空航天大学 Equipment multistage abnormal state identification method and device based on multivariate state estimation technology
CN113268834B (en) * 2021-06-16 2022-08-26 海驹(浙江)智能科技有限公司 Pipeline environment safety and health management system

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104806450A (en) * 2015-03-25 2015-07-29 华北电力大学（保定） Universal gravitation neural network based wind power system MPPT control method

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101872165A (en) * 2010-06-13 2010-10-27 西安交通大学 Method for fault diagnosis of wind turbines on basis of genetic neural network
CN107725283B (en) * 2017-09-19 2019-05-21 江苏方天电力技术有限公司 A kind of fan trouble detection method based on deepness belief network model
CN107977508A (en) * 2017-11-29 2018-05-01 北京优利康达科技股份有限公司 A kind of dynamo bearing failure prediction method
CN108051211B (en) * 2017-12-29 2019-10-18 湖南优利泰克自动化系统有限公司 A kind of wind generator set main shaft holds temperature pre-warning diagnostic method
CN108680358A (en) * 2018-03-23 2018-10-19 河海大学 A kind of Wind turbines failure prediction method based on bearing temperature model

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104806450A (en) * 2015-03-25 2015-07-29 华北电力大学（保定） Universal gravitation neural network based wind power system MPPT control method

Also Published As

Publication number Publication date
CN110108457A (en) 2019-08-09

Similar Documents

Publication Publication Date Title
JP7069269B2 (en) Semi-supervised methods and systems for deep anomaly detection for large industrial surveillance systems based on time series data using digital twin simulation data
CN112202736B (en) Communication network anomaly classification method based on statistical learning and deep learning
Nasiri et al. Fracture mechanics and mechanical fault detection by artificial intelligence methods: A review
CN107725283B (en) A kind of fan trouble detection method based on deepness belief network model
CN106682814B (en) Wind turbine generator fault intelligent diagnosis method based on fault knowledge base
CN105095963B (en) A kind of Precise Diagnosis and the method for prediction wind-tunnel facilities failure
Jia et al. A review of PHM Data Competitions from 2008 to 2017: Methodologies and Analytics
CN107844799B (en) Water chilling unit fault diagnosis method of integrated SVM (support vector machine) mechanism
CN103983453B (en) A kind of executing agency of aero-engine and the differentiating method of sensor fault diagnosis
CN102393882B (en) Method for monitoring and diagnosing indoor air quality (IAQ) sensor on line
CN110006552B (en) Method for detecting abnormal temperature of equipment
CN110414154B (en) Fan component temperature abnormity detection and alarm method with double measuring points
CN112067335A (en) Power plant blower fault early warning method based on multivariate state estimation
CN110108457B (en) Primary fan shaft temperature diagnosis method based on universal gravitation neural network
CN108921230A (en) Method for diagnosing faults based on class mean value core pivot element analysis and BP neural network
CN109675935A (en) A kind of IPCA operation of rolling on-line fault diagnosis method becoming control limit
CN116292367B (en) Power plant fan system abnormal condition detection method based on one-dimensional convolution
CN108257365A (en) A kind of industrial alarm designs method based on global nonspecific evidence dynamic fusion
CN110986407A (en) Fault diagnosis method for centrifugal water chilling unit
CN114326486B (en) Process monitoring method based on probability slow feature analysis and elastic weight consolidation
CN106354125A (en) Method for utilizing block PCA (Principal Component Analysis) to detect fault of chemical process
CN108416386A (en) A kind of method and system judged extremely for Hydropower Unit bearing temperature
CN114936520A (en) Steam turbine fault early warning analysis method based on improved MSET
CN112763215B (en) Multi-working-condition online fault diagnosis method based on modular federal deep learning
CN113187674A (en) Fault determination method and system for wind turbine generator pitch system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
TA01 Transfer of patent application right
TA01 Transfer of patent application right

Effective date of registration: 20190830

Address after: 130012 Changchun high tech Zone, Jilin province Wei Shan Road, No. 3195

Applicant after: Datang northeast electric power test and Research Institute Co., Ltd.

Applicant after: Harbin No.1 Thermoelectric Plant of Datang Heilongjiang Power Generation Co., Ltd.

Address before: 130012 Changchun high tech Zone, Jilin province Wei Shan Road, No. 3195

Applicant before: Datang northeast electric power test and Research Institute Co., Ltd.

CB03 Change of inventor or designer information
CB03 Change of inventor or designer information

Inventor after: Ma Boyang

Inventor after: Li Jintuo

Inventor after: Qu Guangshun

Inventor after: Ma Wenyuan

Inventor after: Li Qiang

Inventor after: Dong Wei

Inventor after: Wang Wentie

Inventor after: Qi Bing

Inventor before: Ma Boyang

Inventor before: Li Jintuo

Inventor before: Qu Guangshun

Inventor before: Ma Wenyuan

Inventor before: Li Qiang

Inventor before: Dong Wei

Inventor before: Wang Wentie

Inventor before: Qi Bing

SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant