CN110794683A - Wind power gear box state evaluation method based on deep neural network and kurtosis characteristics - Google Patents
Wind power gear box state evaluation method based on deep neural network and kurtosis characteristics Download PDFInfo
- Publication number
- CN110794683A CN110794683A CN201911182129.8A CN201911182129A CN110794683A CN 110794683 A CN110794683 A CN 110794683A CN 201911182129 A CN201911182129 A CN 201911182129A CN 110794683 A CN110794683 A CN 110794683A
- Authority
- CN
- China
- Prior art keywords
- state
- gearbox
- kurtosis
- gear box
- neural network
- 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.)
- Withdrawn
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Wind Motors (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention relates to a wind power gear box state evaluation method based on a deep neural network and kurtosis characteristics, which adopts the kurtosis characteristics of vibration signals acquired by a vibration sensor as input values reflecting the vibration characteristics of a gear box, inputs a state matrix D consisting of wind power unit state data to be subjected to state evaluation into a deep neural network input layer, predicts the state of the gear box, further performs state evaluation on the gear box, and facilitates quantitative state detection of the gear box; by adding the vibration sensor and taking a 10min kurtosis characteristic value as model input, not only can the data monitored by an SCADA system of the wind turbine generator be perfected, but also the abnormality of a gear box of the wind turbine generator can be more easily reflected in the operation data; the deep neural network is used for monitoring the wind turbine gearbox in real time, accuracy is high, calculation time is fast, the deep neural network can be uniformly deployed on the wind turbines of the same type in the same wind field, and operation safety of the wind turbines is guaranteed.
Description
Technical Field
The invention relates to a wind turbine generator gearbox state evaluation method based on a deep neural network and kurtosis characteristics.
Background
The large-scale wind turbine generator is used as outdoor power generation equipment, and the running state of the large-scale wind turbine generator is influenced by natural environments such as wind speed, wind direction, temperature and the like and is often severe. The gear box is used as key equipment of the wind turbine generator, bears variable speed and variable load and severe working environment, and once the gear box breaks down, the wind turbine generator is long in downtime and large in power generation loss. If the Data of the SCADA (supervisory Control And Data acquisition) system of the wind turbine generator set can be monitored And analyzed by using a state detection technology, the running state of the gearbox of the wind turbine generator set is evaluated, abnormal signs of the gearbox are found in advance, the fault time of the gearbox can be effectively reduced, And the reliability of the wind turbine generator set And the economic benefit of a wind power plant are improved. Therefore, if the state of the gearbox can be judged in advance in the early stage of the failure of the gearbox of the wind turbine generator, the failure early warning is sent out, and the failure early warning is important for the operation and production of the wind power plant.
Disclosure of Invention
In view of the above situation, in order to overcome the defects of the prior art, the invention aims to provide a wind turbine generator gearbox state evaluation method based on a deep neural network and kurtosis characteristics, which can effectively solve the problem of wind turbine gearbox state evaluation.
The technical scheme of the invention is as follows:
a wind power gear box state evaluation method based on a deep neural network and kurtosis characteristics comprises the following steps:
step 1: vibration sensors are additionally arranged on a low-speed shaft, a medium-speed shaft and a high-speed shaft of a gear box of the wind power generation set, and kurtosis characteristics of vibration signals acquired by the vibration sensors are used as input values for reflecting vibration characteristics of the gear box;
step 2: determining the model of the gearbox to be subjected to state evaluation, the accumulated running time, the current running state information and the historical fault times:
a. the model of the gear box: determining through equipment file query;
b. and (4) accumulating the running time: the accumulated running time of the equipment can be obtained by calling an SCADA system of a wind field;
c. the current running state is as follows: measuring equipment on site to obtain state information of vibration characteristics and temperature characteristics of the equipment, comparing the measured data with data called by an SCADA system, and correcting the data of the SCADA system to ensure that the running state data is correct by taking the measured data as the standard when the measured data and the data are different;
d. the historical failure times are as follows: historical fault information of the wind turbine generator is inquired through an SCADA system, and historical fault times are determined;
and step 3: recording the average value of the unit states in one time period every 10min through an SCADA system to form operation data, classifying and summarizing all wind turbine units with the same type of faults, and intensively calling the operation data of the wind turbine units with the faults within 0-48 h before the faults for collection to form a fault data set; collecting the operation data of the wind turbine generators which normally operate in the same model of the wind field to form a normal data set;
the operation data comprises: wind speed-v, power (generated power) -P, gear box lubricating oil temperature-Tgtemperature-T of main bearing of gear boxbAmbient temperature-TeLow speed shaft radial vibration signal kurtosis-V1And the radial vibration signal kurtosis of the medium speed shaft is-V2High speed shaft radial vibration signal kurtosis-V3And the kurtosis of vibration signals at the gear ring of the lower box body is-V4；
And 4, step 4: establishing a wind power gear box state data set
a. Aiming at a normal data set and a fault data set, parameters in the operating data are used as the composition of an input matrix, and different modes are selected for normalization according to the types of the parameters:
1) the condition of the gearbox is better when the parameter value is smaller
When the parameters are the temperature of lubricating oil of the gearbox, the temperature of a main bearing of the gearbox, the kurtosis of a radial vibration signal of a low-speed shaft, the kurtosis of a radial vibration signal of a medium-speed shaft, the kurtosis of a radial vibration signal of a high-speed shaft and the kurtosis of a vibration signal at a gear ring of a low box body, the smaller the numerical value is, the more stable and healthy the operation of the gearbox is, and the formula for calculating the degradation degree is as follows:
in the formula, x is the obtained parameter, xminIs the minimum value, x, of the parameter in normal operation of the unitmaxThe maximum value of the parameter when the unit normally operates;
2) the state of the gear box is better when the parameter value is within a certain range
When the parameters are environment temperature, power and wind speed, the state of the gear box in a certain area operates optimally, and the formula for calculating the degradation degree is as follows:
in the formula, x is the obtained parameter, xminIs the minimum value, x, of the parameter in normal operation of the unitminFor the maximum value of this parameter during normal operation of the unit, [ x ]a,xb]Is the optimum range for the parameter;
b. forming a matrix containing gearbox state information
For a certain time i, the state of the gearbox can be normalized by state parameters to form a vector X (i) as shown in formula 1:
X(i)＝[v,P,Tg,Te,T,V1,V2,V3,V4]T(1)
in the formula, the corresponding relationship between the parameter name and the parameter symbol is: wind speed-v, power-P, gear box lubricating oil temperature-Tgtemperature-T of main bearing of gear boxbAmbient temperature-TeLow speed shaft radial vibration signal kurtosis-V1And the radial vibration signal kurtosis of the medium speed shaft is-V2High speed shaft radial vibration signal kurtosis-V3And the kurtosis of vibration signals at the gear ring of the lower box body is-V4；
Because the state of the gearbox is a developing process, in order to enable the convolutional neural network to comprehensively grasp the state of the gearbox, m historical observation vectors X (i) form a state matrix D shown in a formula 2:
D＝[X(n+1),X(n+2),X(n+3)...X(n+m)](2)
the state matrix D is composed of m gearbox state vectors from the moment n to the moment n + m, and the matrix D can represent the dynamic process of the equipment in the m moments;
finally forming a plurality of normal state matrixes and a plurality of fault state matrixes;
and 5: training of deep neural networks
Training a network under a Caffe framework, wherein a multilayer deep neural network structure with a convolution structure is used in the training, a plurality of normal state matrixes and fault state matrixes are led in, a training and testing data set of the deep neural network is established, the deep neural network is formed, an input layer of the deep neural network is a state matrix D of a gearbox of a wind turbine generator, and an output layer of the deep neural network is the probability that the gearbox is in a fault state;
step 6: wind turbine generator gearbox state assessment
Inputting a state matrix D consisting of wind turbine state data to be subjected to state evaluation into a deep neural network input layer, predicting the state of a gearbox, further performing state evaluation on the gearbox, wherein the output evaluation result is a parameter k between 0 and 1, the parameter k is the probability that the gearbox output by the deep neural network is in a fault state, judging that the gearbox is likely to have faults when k is larger than or equal to 0.5, and judging that the gearbox can stably run when k is smaller than 0.5.
Compared with the prior art, the method has the following beneficial technical effects:
(1) the method has the advantages that a parameter k between 0 and 1 representing the state of the gearbox can be output in real time, when the k is more than or equal to 0.5, the possibility that the gearbox is about to break down is judged, when the k is less than 0.5, the gearbox can be judged to run stably, and quantitative state detection of the gearbox is facilitated;
(2) by adding the vibration sensor and taking a 10min kurtosis characteristic value as model input, not only can the data monitored by an SCADA system of the wind turbine generator be perfected, but also the abnormality of a gear box of the wind turbine generator can be more easily reflected in the operation data;
(3) the deep neural network is used for monitoring the wind turbine gearbox in real time, accuracy is high, calculation time is fast, the deep neural network can be uniformly deployed on the wind turbines of the same type in the same wind field, and operation safety of the wind turbines is guaranteed.
Drawings
FIG. 1 is a schematic diagram of probability values of test samples in a test, which are output by the method of the present invention and correspond to the identified fault states.
Detailed Description
The following examples further illustrate the embodiments of the present invention in detail.
A wind power gear box state evaluation method based on a deep neural network and kurtosis characteristics comprises the following steps:
step 1: the wind field wind turbine generator gearbox is characterized in that vibration sensors are additionally arranged on a low-speed shaft, a medium-speed shaft and a high-speed shaft of the wind field wind turbine generator gearbox, and meanwhile, the fact that vibration signals change along with various factors such as wind speed and wind direction is considered, so that kurtosis characteristics of the vibration signals collected by the vibration sensors are used as input values for reflecting vibration characteristics of the gearbox;
because the vibration sensor and the SCADA system have larger frequency difference, the kurtosis characteristic value is calculated every 1s for the obtained vibration signal, and the average value of the larger first 30 percent of the characteristic value within 10min is taken as the characteristic variable input;
the kurtosis of an oscillatory signal is a time domain parameter of dimension one, defined as the normalized 4 th order central moment of the oscillatory signal, which is mathematically described as:
wherein: x is the number ofqIs the kurtosis; x is the number ofiThe vibration signal value at the moment i;
the average value of the vibration signals is obtained; n is the signal length, σ4Is the vibration signal variance.
The kurtosis index can well reflect the impact components in the vibration signal, and the more the impact, the larger the kurtosis.
Step 2: determining the model of the gearbox to be state evaluated, the accumulated running time (hours), the current running state information, the historical failure times:
a. the model of the gear box: determining through equipment file query;
b. and (4) accumulating the running time: the accumulated running time of the equipment can be obtained by calling an SCADA system of a wind field;
c. the current running state is as follows: measuring equipment on site to obtain state information of vibration characteristics and temperature characteristics of the equipment, comparing the measured data with data called by an SCADA system, and correcting the data of the SCADA system to ensure that the running state data is correct by taking the measured data as the standard when the measured data and the data are different;
d. the historical failure times are as follows: historical fault information of the wind turbine generator is inquired through an SCADA system, and historical fault times are determined;
and step 3: recording the average value of the state of the wind turbine generators in a period of once every 10min through an SCADA system, classifying and summarizing all the wind turbine generators with the same type of faults, and intensively calling and collecting the operation data of the wind turbine generators with the faults within 0-48 h before the faults to form a fault data set; collecting the operation data of the wind turbine generators which normally operate in the same model of the wind field to form a normal data set;
the operation data comprises: wind speed-v, power (generated power) -P, gear box lubricating oil temperature-Tgtemperature-T of main bearing of gear boxbAmbient temperature-TeLow speed shaft radial vibration signal kurtosis-V1And the radial vibration signal kurtosis of the medium speed shaft is-V2High speed shaft radial vibration signal kurtosis-V3And the kurtosis of vibration signals at the gear ring of the lower box body is-V4；
And 4, step 4: establishing a wind power gear box state data set
a. Aiming at a normal data set and a fault data set, parameters in the operating data are used as the composition of an input matrix, and different modes are selected for normalization according to the types of the parameters:
1) the condition of the gearbox is better when the parameter value is smaller
When the parameters are the temperature of lubricating oil of the gearbox, the temperature of a main bearing of the gearbox, the kurtosis of a radial vibration signal of a low-speed shaft, the kurtosis of a radial vibration signal of a medium-speed shaft, the kurtosis of a radial vibration signal of a high-speed shaft and the kurtosis of a vibration signal at a gear ring of a low box body, the smaller the numerical value is, the more stable and healthy the operation of the gearbox is, and the formula for calculating the degradation degree is as follows:
in the formula, x is the obtained parameter, xminIs the minimum value, x, of the parameter in normal operation of the unitmaxThe maximum value of the parameter when the unit normally operates;
2) the state of the gear box is better when the parameter value is within a certain range
When the parameters are environment temperature, power and wind speed, the state of the gear box in a certain area operates optimally, and the formula for calculating the degradation degree is as follows:
in the formula, x is the obtained parameter, xminIs the minimum value, x, of the parameter in normal operation of the unitminFor the maximum value of this parameter during normal operation of the unit, [ x ]a,xb]For the optimal range of the parameter, the optimal range can be determined according to the corresponding model query 'wind turbine generator operation specification';
b. forming a matrix containing gearbox state information
For a certain time i, the state of the gearbox can be normalized by state parameters to form a vector X (i) as shown in formula 1:
X(i)＝[v,P,Tg,Te,T,V1,V2,V3,V4]T(1)
in the formula, the corresponding relationship between the parameter name and the parameter symbol is: wind speed-v, power-P, gear box lubricating oil temperature-Tgtemperature-T of main bearing of gear boxbAmbient temperature-TeLow speed shaft radial vibration signal kurtosis-V1And the radial vibration signal kurtosis of the medium speed shaft is-V2High speed shaft radial vibration signal kurtosis-V3And the kurtosis of vibration signals at the gear ring of the lower box body is-V4；
Since the state of the gearbox is a developing process, in order to enable a Convolutional Neural Network (CNN) to comprehensively grasp the state of the gearbox, m historical observation vectors x (i) are formed into a state matrix D shown in formula 2:
D＝[X(n+1),X(n+2),X(n+3)...X(n+m)](2)
the state matrix D is composed of m gearbox state vectors from the moment n to the moment n + m, and the matrix D can represent the dynamic process of the equipment in the m moments;
finally forming a plurality of normal state matrixes and a plurality of fault state matrixes;
and 5: training of deep neural networks
Training a network under a Caffe framework, introducing a plurality of normal state matrixes and fault state matrixes by using a multilayer deep neural network structure with a convolution structure in training, establishing a training and testing data set of the deep neural network, and searching an optimal network structure through trial and error; the deep neural network obtains higher accuracy by adjusting network parameters; establishing a deep neural network with 4 convolutional layers, 3 pooling layers and 1 full-connection layer, wherein the input layer is a state matrix D of a gearbox of a wind turbine generator, and the output layer is the probability that the gearbox is in a fault state;
during specific operation, a plurality of normal state matrixes and fault state matrixes D are respectively used as samples and stored in a 'CSV' file, normal and fault labels are added, the 'CSV' file is read through a Caffe program to establish a training and testing data set of a deep neural network, and optimal network structure parameters are searched through trial and error; the deep neural network obtains higher accuracy by adjusting network parameters;
step 6: wind turbine generator gearbox state assessment
Inputting a state matrix D consisting of wind turbine state data to be subjected to state evaluation into a deep neural network input layer, predicting the state of a gearbox, further performing state evaluation on the gearbox, wherein the output evaluation result is a parameter k between 0 and 1, the parameter k is the probability that the gearbox output by the deep neural network is in a fault state, judging that the gearbox is likely to have faults when k is larger than or equal to 0.5, and judging that the gearbox can stably run when k is smaller than 0.5.
The invention obtains better results through tests on different units of the same model of the wind power plant. The contents of the test were as follows:
a certain number of units of a certain wind field are taken as test samples, operation data of the units in a certain period are extracted as input, and the state of the wind turbine gearbox samples is shown in the following table:
wind turbine gearbox sample state
And (3) forming a real-time state matrix D of the gearbox by using kurtosis characteristic values measured by an SCADA system and a vibration sensor of the unit.
Establishing a gearbox state matrix rule according to the method, selecting a gearbox state matrix which is 10min to 4h before the fault occurs at the last moment and a normally running gearbox state matrix, identifying by using the deep neural network model established in the step 5 of the method, and outputting a probability value corresponding to the identified fault state as shown in figure 1.
As can be seen from the figure, when the time is advanced from the fault occurrence time, the probability that the model judges the gearbox state to be the fault is all below 0.5, so that the gearbox is identified to be in the normal state, and as the time of the state matrix is closer to the fault time, the probability that the model judges the gearbox to be in the fault state begins to increase and finally exceeds 0.5, so that the model judges the gearbox to be in the fault state. At the same time, it can be found that gear wear faults are identified earlier than gearbox oil supply faults. In the field inspection, the oil supply failure is found to be the overload of the oil pump caused by the high viscosity of the gear lubricating oil, which is closely related to the ambient temperature at the time, so that the oil supply failure has certain abruptness. Subsequent gear wear failures can be more pronounced from vibration signature and other signatures, and thus a determination of a failure condition can be made earlier. Therefore, the model can be demonstrated to carry out fault early warning and state monitoring on the wind turbine generator of the wind field to a certain extent, and the phenomenon that the generator is more damaged due to the fault of the gear box is avoided.
Claims (1)
1. A wind power gear box state evaluation method based on a deep neural network and kurtosis characteristics is characterized by comprising the following steps:
step 1: vibration sensors are additionally arranged on a low-speed shaft, a medium-speed shaft and a high-speed shaft of a gear box of the wind power generation set, and kurtosis characteristics of vibration signals acquired by the vibration sensors are used as input values for reflecting vibration characteristics of the gear box;
step 2: determining the model of the gearbox to be subjected to state evaluation, the accumulated running time, the current running state information and the historical fault times:
a. the model of the gear box: determining through equipment file query;
b. and (4) accumulating the running time: the accumulated running time of the equipment can be obtained by calling an SCADA system of a wind field;
c. the current running state is as follows: measuring equipment on site to obtain state information of vibration characteristics and temperature characteristics of the equipment, comparing the measured data with data called by an SCADA system, and correcting the data of the SCADA system to ensure that the running state data is correct by taking the measured data as the standard when the measured data and the data are different;
d. the historical failure times are as follows: historical fault information of the wind turbine generator is inquired through an SCADA system, and historical fault times are determined;
and step 3: recording the average value of the unit states in one time period every 10min through an SCADA system to form operation data, classifying and summarizing all wind turbine units with the same type of faults, and intensively calling the operation data of the wind turbine units with the faults within 0-48 h before the faults for collection to form a fault data set; collecting the operation data of the wind turbine generators which normally operate in the same model of the wind field to form a normal data set;
the operation data comprises: wind speed-v, power (generated power) -P, gear box lubricating oil temperature-Tgtemperature-T of main bearing of gear boxbAmbient temperature-TeLow speed shaft radial vibration signal kurtosis-V1And the radial vibration signal kurtosis of the medium speed shaft is-V2High speed shaft radial vibration signal kurtosis-V3And the kurtosis of vibration signals at the gear ring of the lower box body is-V4；
And 4, step 4: establishing a wind power gear box state data set
a. Aiming at a normal data set and a fault data set, parameters in the operating data are used as the composition of an input matrix, and different modes are selected for normalization according to the types of the parameters:
1) the condition of the gearbox is better when the parameter value is smaller
When the parameters are the temperature of lubricating oil of the gearbox, the temperature of a main bearing of the gearbox, the kurtosis of a radial vibration signal of a low-speed shaft, the kurtosis of a radial vibration signal of a medium-speed shaft, the kurtosis of a radial vibration signal of a high-speed shaft and the kurtosis of a vibration signal at a gear ring of a low box body, the smaller the numerical value is, the more stable and healthy the operation of the gearbox is, and the formula for calculating the degradation degree is as follows:
in the formula, x is the obtained parameter, xminIs the minimum value, x, of the parameter in normal operation of the unitmaxThe maximum value of the parameter when the unit normally operates;
2) the state of the gear box is better when the parameter value is within a certain range
When the parameters are environment temperature, power and wind speed, the state of the gear box in a certain area operates optimally, and the formula for calculating the degradation degree is as follows:
in the formula, x is the obtained parameter, xminIs the minimum value, x, of the parameter in normal operation of the unitminFor the maximum value of this parameter during normal operation of the unit, [ x ]a,xb]Is the optimum range for the parameter;
b. forming a matrix containing gearbox state information
For a certain time i, the state of the gearbox can be normalized by state parameters to form a vector X (i) as shown in formula 1:
X(i)＝[v,P,Tg,Te,T,V1,V2,V3,V4]T(1)
in the formula, the corresponding relationship between the parameter name and the parameter symbol is: wind speed-v, power-P, gear box lubricating oil temperature-Tgtemperature-T of main bearing of gear boxbAmbient temperature-TeLow speed shaft radial vibration signal kurtosis-V1And the radial vibration signal kurtosis of the medium speed shaft is-V2High speed shaft radial vibration signal kurtosis-V3And the kurtosis of vibration signals at the gear ring of the lower box body is-V4；
Because the state of the gearbox is a developing process, in order to enable the convolutional neural network to comprehensively grasp the state of the gearbox, m historical observation vectors X (i) form a state matrix D shown in a formula 2:
D＝[X(n+1),X(n+2),X(n+3)...X(n+m)](2)
the state matrix D is composed of m gearbox state vectors from the moment n to the moment n + m, and the matrix D can represent the dynamic process of the equipment in the m moments;
finally forming a plurality of normal state matrixes and a plurality of fault state matrixes;
and 5: training of deep neural networks
Training a network under a Caffe framework, wherein a multilayer deep neural network structure with a convolution structure is used in the training, a plurality of normal state matrixes and fault state matrixes are led in, a training and testing data set of the deep neural network is established, the deep neural network is formed, an input layer of the deep neural network is a state matrix D of a gearbox of a wind turbine generator, and an output layer of the deep neural network is the probability that the gearbox is in a fault state;
step 6: wind turbine generator gearbox state assessment
Inputting a state matrix D consisting of wind turbine state data to be subjected to state evaluation into a deep neural network input layer, predicting the state of a gearbox, further performing state evaluation on the gearbox, wherein the output evaluation result is a parameter k between 0 and 1, the parameter k is the probability that the gearbox output by the deep neural network is in a fault state, judging that the gearbox is likely to have faults when k is larger than or equal to 0.5, and judging that the gearbox can stably run when k is smaller than 0.5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911182129.8A CN110794683A (en) | 2019-11-27 | 2019-11-27 | Wind power gear box state evaluation method based on deep neural network and kurtosis characteristics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911182129.8A CN110794683A (en) | 2019-11-27 | 2019-11-27 | Wind power gear box state evaluation method based on deep neural network and kurtosis characteristics |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110794683A true CN110794683A (en) | 2020-02-14 |
Family
ID=69446351
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911182129.8A Withdrawn CN110794683A (en) | 2019-11-27 | 2019-11-27 | Wind power gear box state evaluation method based on deep neural network and kurtosis characteristics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110794683A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111639842A (en) * | 2020-05-20 | 2020-09-08 | 湖北博华自动化***工程有限公司 | Equipment health evaluation method, evaluation system and equipment health prediction method |
| CN111950505A (en) * | 2020-08-24 | 2020-11-17 | 湖南科技大学 | State evaluation method for wind driven generator sensor of SSA-AANN |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103512751A (en) * | 2013-07-03 | 2014-01-15 | 辽宁大学 | Bearing health state identification method based on probabilistic neural network |
| CN108204895A (en) * | 2017-11-29 | 2018-06-26 | 中国电力科学研究院有限公司 | A kind of Wind turbines change gear box fault diagnosis method and system |
| CN109063308A (en) * | 2018-07-26 | 2018-12-21 | 北京航空航天大学 | A kind of health evaluating method based on depth quantum learning |
| CN109425483A (en) * | 2017-09-04 | 2019-03-05 | 锐电科技有限公司 | Running of wind generating set status assessment and prediction technique based on SCADA and CMS |
| US20190095781A1 (en) * | 2017-09-23 | 2019-03-28 | Nanoprecise Sci Corp. | System and method for automated fault diagnosis and prognosis for rotating equipment |
| CN109580218A (en) * | 2018-12-08 | 2019-04-05 | 上海电力学院 | A kind of state of fan gear box recognition methods based on likelihood learning machine |
-
2019
- 2019-11-27 CN CN201911182129.8A patent/CN110794683A/en not_active Withdrawn
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103512751A (en) * | 2013-07-03 | 2014-01-15 | 辽宁大学 | Bearing health state identification method based on probabilistic neural network |
| CN109425483A (en) * | 2017-09-04 | 2019-03-05 | 锐电科技有限公司 | Running of wind generating set status assessment and prediction technique based on SCADA and CMS |
| US20190095781A1 (en) * | 2017-09-23 | 2019-03-28 | Nanoprecise Sci Corp. | System and method for automated fault diagnosis and prognosis for rotating equipment |
| CN108204895A (en) * | 2017-11-29 | 2018-06-26 | 中国电力科学研究院有限公司 | A kind of Wind turbines change gear box fault diagnosis method and system |
| CN109063308A (en) * | 2018-07-26 | 2018-12-21 | 北京航空航天大学 | A kind of health evaluating method based on depth quantum learning |
| CN109580218A (en) * | 2018-12-08 | 2019-04-05 | 上海电力学院 | A kind of state of fan gear box recognition methods based on likelihood learning machine |
Non-Patent Citations (2)
| Title |
|---|
| 罗毅 等: "基于小波包与PCA遗传神经网络相结合的齿轮箱故障诊断方法", 《化工自动化及仪表》 * |
| 贾子文等: "基于数据挖掘的风电机组齿轮箱运行状态分析", 《中国机械工程》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111639842A (en) * | 2020-05-20 | 2020-09-08 | 湖北博华自动化***工程有限公司 | Equipment health evaluation method, evaluation system and equipment health prediction method |
| CN111950505A (en) * | 2020-08-24 | 2020-11-17 | 湖南科技大学 | State evaluation method for wind driven generator sensor of SSA-AANN |
| CN111950505B (en) * | 2020-08-24 | 2023-08-29 | 湖南科技大学 | SSA-AANN wind driven generator sensor state evaluation method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109524139B (en) | Real-time equipment performance monitoring method based on equipment working condition change | |
| Kim et al. | Use of SCADA data for failure detection in wind turbines | |
| Fischer et al. | Condition monitoring of wind turbines: State of the art, user experience and recommendations | |
| Zaher et al. | Online wind turbine fault detection through automated SCADA data analysis | |
| US6587737B2 (en) | Method for the monitoring of a plant | |
| CA2710902A1 (en) | Integrated condition based maintenance system for wind turbines | |
| CN110529202B (en) | Steam turbine fault diagnosis detection and early warning method and system based on power plant data | |
| CN114444382B (en) | Wind turbine generator system gear box fault diagnosis and analysis method based on machine learning algorithm | |
| CN110907170B (en) | Wind turbine generator gearbox bearing temperature state monitoring and fault diagnosis method | |
| CN103925155A (en) | Self-adaptive detection method for abnormal wind turbine output power | |
| Praveenkumar et al. | Comparison of vibration, sound and motor current signature analysis for detection of gear box faults | |
| CN110794683A (en) | Wind power gear box state evaluation method based on deep neural network and kurtosis characteristics | |
| CN112696481A (en) | Intelligent diagnosis method and device for shaft temperature abnormity of wind turbine generator gearbox | |
| Kim et al. | Design of wind turbine fault detection system based on performance curve | |
| Tchakoua et al. | New trends and future challenges for wind turbines condition monitoring | |
| CN110826600B (en) | Engine surge prediction method based on adaptive resonance network online incremental learning | |
| CN114577470A (en) | Fault diagnosis method and system for fan main bearing | |
| Carroll et al. | Availability improvements from condition monitoring systems and performance based maintenance contracts | |
| Karadayi et al. | Fault-related Alarm Detection of a Wind Turbine SCADA System | |
| US11339763B2 (en) | Method for windmill farm monitoring | |
| McKinnon et al. | Investigation of anomaly detection technique for wind turbine pitch systems | |
| Tutiv'en et al. | Wind turbine main bearing condition monitoring via convolutional autoencoder neural networks | |
| Zhang et al. | Probability warning for wind turbine gearbox incipient faults based on SCADA data | |
| Bilendo et al. | Wind turbine anomaly detection based on SCADA data | |
| CN112699598A (en) | Intelligent diagnosis method and device for abnormal oil temperature of gear box |
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 | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20200214 |
|
| WW01 | Invention patent application withdrawn after publication |