CN114444394A - Gas compressor performance degradation prediction algorithm based on data driving - Google Patents
Gas compressor performance degradation prediction algorithm based on data driving Download PDFInfo
- Publication number
- CN114444394A CN114444394A CN202210102459.7A CN202210102459A CN114444394A CN 114444394 A CN114444394 A CN 114444394A CN 202210102459 A CN202210102459 A CN 202210102459A CN 114444394 A CN114444394 A CN 114444394A
- Authority
- CN
- China
- Prior art keywords
- compressor
- neural network
- data
- efficiency
- model
- 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.)
- Pending
Links
- 230000015556 catabolic process Effects 0.000 title claims abstract description 31
- 238000006731 degradation reaction Methods 0.000 title claims abstract description 31
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 25
- 238000013528 artificial neural network Methods 0.000 claims abstract description 45
- 230000008859 change Effects 0.000 claims abstract description 12
- 238000012360 testing method Methods 0.000 claims abstract description 11
- 230000015654 memory Effects 0.000 claims abstract description 6
- 238000013210 evaluation model Methods 0.000 claims abstract description 4
- 125000004122 cyclic group Chemical group 0.000 claims abstract description 3
- 238000006243 chemical reaction Methods 0.000 claims description 22
- 238000000034 method Methods 0.000 claims description 16
- 238000012549 training Methods 0.000 claims description 12
- 230000007613 environmental effect Effects 0.000 claims description 8
- 238000007781 pre-processing Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 4
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 230000004913 activation Effects 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 2
- 238000004088 simulation Methods 0.000 abstract description 4
- 238000012545 processing Methods 0.000 abstract description 3
- 238000004458 analytical method Methods 0.000 abstract description 2
- 238000005065 mining Methods 0.000 abstract description 2
- 230000000306 recurrent effect Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 4
- 230000007787 long-term memory Effects 0.000 description 4
- 238000002485 combustion reaction Methods 0.000 description 3
- 230000006403 short-term memory Effects 0.000 description 3
- 230000008034 disappearance Effects 0.000 description 2
- 238000004880 explosion Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 238000005406 washing Methods 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 239000000446 fuel Substances 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000002945 steepest descent method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/28—Design optimisation, verification or simulation using fluid dynamics, e.g. using Navier-Stokes equations or computational fluid dynamics [CFD]
-
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- 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
- G06N3/045—Combinations of networks
-
- 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
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/08—Fluids
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Algebra (AREA)
- Fluid Mechanics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Control Of Positive-Displacement Pumps (AREA)
Abstract
The invention provides a compressor performance degradation prediction algorithm based on data driving, which comprises the following steps: s1, establishing a compressor performance change evaluation model; s11, collecting power plant data; s12, taking the efficiency and the flow capacity of the compressor as indexes for evaluating the performance change of the compressor; s13, establishing a thermodynamic model of the compressor; s14, calculating the efficiency and the flow capacity of the compressor key component through the thermodynamic model to obtain theoretical efficiency and flow capacity; s21, establishing a deep neural network; and S22, introducing a cyclic neural network and a long-short term memory network to optimize the deep neural network to obtain a performance degradation prediction neural network. According to the invention, a more accurate compressor performance analysis model is established by analyzing and processing a large amount of data generated by the operation of the compressor and mining the hidden relation between the large data and the performance change, so that the simulation test is closer to the actual operation of a real unit, and the model has a higher engineering practical value.
Description
Technical Field
The invention relates to the technical field of gas compressors, in particular to a gas compressor performance degradation prediction algorithm based on data driving.
Background
The gas turbine is an internal combustion type power machine which drives an impeller to rotate at a high speed by a constantly flowing working medium and converts the energy of fuel into output power. The press machine is responsible for compressing air to compress gas at normal temperature and normal pressure into high-pressure gas, the high-pressure gas is input into the combustion chamber to be combusted, and the formed high-temperature high-pressure gas does work in the turbine. Therefore, the performance of the compressor directly affects the performance of the subsequent combustion chamber and the turbine, and due to the variable operation process and the severe working environment, the performance of the compressor is often reduced due to blade scaling, blade corrosion and the like, so that a model algorithm of the compressor needs to be established and relevant simulation needs to be carried out to analyze and predict the performance reduction of the compressor.
The currently generally established compressor model algorithm is a mechanism model algorithm, but because the characteristic curve of the compressor is generally mastered in foreign equipment manufacturers, the very accurate mechanism model algorithm is difficult to establish, so that the actual working characteristic is not consistent with the working characteristic of the modeling algorithm, and the simulation of the algorithm model has larger deviation from the actual result. With the rapid development of computer technology and data-driven theory, the data-driven modeling algorithm gradually becomes the main trend of modeling the compressor. In summary, an algorithm capable of predicting the performance degradation of the compressor more accurately needs to be invented.
Disclosure of Invention
The invention provides a data-driven compressor performance degradation prediction algorithm, which solves the problem that the existing compressor performance degradation prediction algorithm has larger deviation with the actual result.
In order to achieve the above purposes, the technical scheme adopted by the invention is as follows:
a data-driven compressor performance degradation prediction algorithm, comprising:
s1, establishing a compressor performance change evaluation model;
s11, collecting power plant data;
s12, taking the efficiency and the flow capacity of the compressor as indexes for evaluating the performance change of the compressor;
s13, establishing a thermodynamic model of the compressor;
s14, calculating the efficiency and the flow capacity of the compressor key component through the thermodynamic model to obtain theoretical efficiency and flow capacity;
s15, optimizing the deep neural network by a Levenberg-Marquardt method to obtain an ISO standard state reduced neural network;
s16, establishing an ISO standard state conversion model of the key component through the ISO standard state conversion neural network;
s17, inputting the power plant data into an ISO standard state conversion model, and converting the efficiency and the flow capacity of key components in different running states into values under full load and ISO working conditions through the ISO standard state conversion model;
s18, transversely comparing the values of the key components under full load and ISO working conditions to obtain the efficiency and the flow capacity under the standard condition;
s2, establishing a compressor performance prediction model;
s21, establishing a deep neural network;
s22, introducing a cyclic neural network and a long-short term memory network to optimize the deep neural network to obtain a performance degradation prediction neural network;
and S23, inputting the power plant data collected in the S11 into a performance degradation prediction neural network to obtain the efficiency and the flow capacity of the compressor for predicting degradation.
Preferably, the compressor thermodynamic model comprises: compressor efficiency, compressor power consumption and compressor flow capacity;
the compressor efficiency formula is:
wherein m isaIs the flow rate of the air compressor,is the specific enthalpy of air at the inlet temperature of the compressor,is the specific enthalpy, W, of air at the isentropic outlet temperature of the compressorc-caConsuming power for the press;
the formula of the power consumption of the compressor is as follows:
wherein m iscaThe air-extracting flow of the air compressor is the air-extracting flow,is the specific enthalpy of air at the outlet temperature of the compressor,the specific enthalpy of air at the average temperature of air extraction of the compressor is obtained;
the formula of the flow capacity of the compressor is as follows:
wherein, T1 refReference value, P, for compressor inlet temperature1 refIs a reference value of the compressor inlet pressure.
Preferably, the ISO standard posture conversion model has three hidden layers, and each hidden layer comprises 256 units.
Preferably, the power plant data includes environmental parameters and performance parameters, the environmental parameters include environmental temperature, pressure and relative humidity, the performance parameters include power, air flow, pressure ratio, compressor inlet temperature and pressure, inlet guide vane angle IGV, compressor outlet temperature and pressure of the gas turbine, and the output parameters of the ISO standard state conversion model include efficiency and flow capacity of the compressor and the turbine.
Preferably, the optimizing the deep neural network comprises the following steps:
collecting power plant data;
preprocessing collected power plant data, wherein the preprocessing comprises the steps of filtering unsteady data of the power plant data through a sliding window method, then removing abnormal points through a Layida criterion, and finally according to the following steps of 7: and 3, randomly dividing the data into a training set and a testing set, and finally obtaining the optimized deep neural network through statistical significance difference test.
Preferably, the compressor performance prediction model has the following formula:
h(t)=φ(Ux(t)+Wh(t-1)+b)
o(t)=Vh(t)+c
wherein x is input, h is hidden unit, o is output, phi is activation function, V, W, U is weight; b. c is the deviation.
The invention has the beneficial effects that:
in the invention, a model-based modeling method has many defects, and a more accurate compressor performance analysis model is established by analyzing and processing a large amount of data generated by the operation of the compressor and mining the hidden relation between the big data and the performance change, so that the simulation test is closer to the actual operation of a real unit, and the model has higher engineering practical value;
the Levenberg-Marquardt method is adopted, so that the convergence speed and the convergence precision of the model are effectively improved, and the training time and the training error are reduced;
the invention adopts the recurrent neural network, and the recurrent neural network can trace a long time period, thereby increasing the accuracy of prediction;
the invention introduces the long-term and short-term memory network to improve the problem of gradient disappearance or explosion of the recurrent neural network and improve the prediction rate of the model.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flow chart of the method of the present invention.
FIG. 2 is a diagram of the deep neural network architecture of the present invention.
FIG. 3a is a graph of the efficiency of the compressor of the present invention after conversion to ISO standard; fig. 3b is a flow energy diagram of the compressor of the invention after conversion to the ISO standard state.
FIG. 4 is a diagram of a recurrent neural network architecture in accordance with the present invention.
FIG. 5 is a diagram of the key blocks of the long term and short term memory of the present invention.
FIG. 6a is a graph comparing the efficiency degradation trend and the prediction result of the compressor of the present invention; FIG. 6b is a graph comparing the degradation trend of the compressor flow capacity and the predicted result.
Detailed Description
It should be noted that the embodiments and features of the embodiments may be combined with each other without conflict. The present invention will be described in detail below with reference to the embodiments with reference to the attached drawings.
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the invention. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
The relative arrangement of the components and steps, the numerical expressions and numerical values set forth in these embodiments do not limit the scope of the present invention unless specifically stated otherwise. Meanwhile, it should be understood that the sizes of the respective portions shown in the drawings are not drawn in an actual proportional relationship for the convenience of description. Techniques, methods, and apparatus known to those of ordinary skill in the relevant art may not be discussed in detail but are intended to be part of the specification where appropriate. In all examples shown and discussed herein, any particular value should be construed as merely illustrative, and not limiting. Thus, other examples of the exemplary embodiments may have different values. It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, further discussion thereof is not required in subsequent figures.
The invention provides a technical scheme that: a data-driven compressor performance degradation prediction algorithm, as shown in fig. 1, includes:
s1, establishing a compressor performance change evaluation model;
the degradation of the compressor generally causes the reduction of the flow capacity and efficiency of the compressor, therefore, the efficiency and the flow capacity of the compressor are used as indexes for evaluating the performance change of the compressor, and the efficiency and the flow capacity of key components are calculated through a thermodynamic model of the compressor. However, the efficiency and the flow capacity of the compressor are often influenced by different working states, so that an ISO standard state conversion model of a key part is established through an optimized deep neural network, and the efficiency and the flow capacity in different running states are converted into values under full load and ISO working conditions and then transversely compared to evaluate the performance of the compressor.
S11, collecting power plant data;
s12, taking the efficiency and the flow capacity of the compressor as indexes for evaluating the performance change of the compressor;
s13, establishing a thermodynamic model of the compressor;
the calculation formula of the key parameters of the compressor is as follows:
compressor efficiency:
wherein m isaThe flow rate of the air compressor is used,andspecific enthalpy, W, of air at compressor inlet temperature and isentropic outlet temperature, respectivelyc-caConsuming power for the press.
Compressor power consumption:
wherein m iscaThe air-extracting flow of the air compressor is the air-extracting flow,andthe air specific enthalpy is respectively the outlet temperature of the compressor and the air extraction average temperature.
Compressor flow capacity:
wherein, T1 refAnd P1 refReference values for compressor inlet temperature and pressure, respectively.
S14, calculating the efficiency and the flow capacity of the compressor key component through the thermodynamic model to obtain theoretical efficiency and flow capacity;
s15, optimizing the deep neural network by a Levenberg-Marquardt method to obtain an ISO standard state reduced neural network;
s16, establishing an ISO standard state conversion model of the key component through the ISO standard state conversion neural network;
the deep neural network model of ISO standard state conversion is shown in FIG. 2, and has three hidden layers, each layer having 256 units. The input parameters are divided into two categories: environmental parameters and performance parameters. The environmental parameters include ambient temperature, pressure and relative humidity. The performance parameters include power, air flow, pressure ratio, compressor inlet temperature and pressure, inlet guide vane angle IGV, and compressor discharge temperature and pressure of the gas turbine. The output parameters are the efficiency and flow capacity of the compressor and the turbine. The back propagation algorithm of the deep neural network is optimized by using a Levenberg-Marquardt method, the Levenberg-Marquardt method is the combination of a gradient steepest descent method and a Newton method, the convergence speed and the convergence precision of the model can be effectively improved, and the training time and the training error are reduced.
The collected power plant data need to be subjected to data preprocessing, unsteady data are filtered by a sliding window method, then abnormal points are removed by a Layouda criterion, and finally, the data preprocessing is carried out according to the following steps of: 3, randomly dividing the training set and the testing set, and finding no significant difference between the two groups after statistical significant difference test. And training the optimized deep neural network model, the unoptimized model and other common algorithms through a training set respectively, testing through a testing set, and comparing the root mean square error, the average absolute error and the R-square of the training result respectively. The mean absolute error and the root mean square error of the optimized model are far smaller than those of other traditional algorithms, and the R square is closer to 1, which shows that the optimized model has higher precision and better fitting degree without overfitting.
The data was sorted chronologically, setting 200 data points at a time as a data set. And training the model by using the data set, and substituting full-load operation data under the ISO working condition into the model to calculate the efficiency and the flow capacity after ISO standard state conversion. And continuously replacing the data in the data set to realize ISO standard state conversion of all the data. Fig. 3a and 3b show the trend of efficiency and flow capacity. The black dotted line represents an off-line water washing node, and it can be seen that the efficiency and the flow capacity of the compressor gradually decrease with the increase of the operation time, and the off-line water washing node is obviously improved and is consistent with the expectation, which indicates that the performance change evaluation method is feasible and effective.
S17, inputting the power plant data into an ISO standard state conversion model, and converting the efficiency and the flow capacity of key components in different running states into values under full load and ISO working conditions through the ISO standard state conversion model;
s18, transversely comparing the values of the key components under full load and ISO working conditions to obtain the efficiency and the flow capacity under the standard condition;
s2, establishing a compressor performance prediction model;
performance prediction is based on past and current operating data, but a general artificial neural network or other traditional methods are difficult to trace back for a long time period, so that the accuracy of prediction is influenced. The recurrent neural network is the most suitable neural network for processing time series data, and is widely applied to the field of life prediction and fault diagnosis, and the architecture is shown in fig. 4, wherein x is input, h is a hidden unit, o is output, L is a loss function, and y is a label of a training set. The superscript t of these elements represents the state at time t, the output at time t being as follows:
h(t)=φ(Ux(t)+Wh(t-1)+b) (4)
o(t)=Vh(t)+c (5)
wherein phi is an activation function; v, W, U is a weight; b. c is the deviation.
S21, establishing a deep neural network;
s22, introducing a circulating neural network and a long-short term memory network to optimize the deep neural network to obtain a performance degradation prediction neural network;
and S23, inputting the power plant data collected in the S11 into a performance degradation prediction neural network to obtain the efficiency and the flow capacity of the compressor for predicting degradation.
However, in error back propagation across multiple time steps, the recurrent neural network has the problem of gradient disappearance or explosion. Therefore, a long-short term memory network is introduced to improve the situation, and the whole structure is the same as that of the recurrent neural network, but the repeated modules are changed, and the modules are shown in fig. 5. In contrast to the recurrent neural network, long-short term memory flows over time, as well as h, and cell state C, which represents long-term memory.
The degradation trend is predicted by long and short term memory, and fig. 6a and 6b are comparisons of the degradation trend and the prediction result in the test set. The error percentages of the efficiency and the flow capacity of the compressor are 0.036% and 0.11%, which shows that the model has higher prediction accuracy and can realize accurate prediction of the performance degradation of the compressor.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be considered to be within the technical scope of the present invention, and the technical solutions and the inventive concepts thereof according to the present invention should be equivalent or changed within the scope of the present invention.
Claims (6)
1. A data-driven compressor performance degradation prediction algorithm, comprising:
s1, establishing a compressor performance change evaluation model;
s11, collecting power plant data;
s12, taking the efficiency and the flow capacity of the compressor as indexes for evaluating the performance change of the compressor;
s13, establishing a thermodynamic model of the compressor;
s14, calculating the efficiency and the flow capacity of the compressor key component through the thermodynamic model to obtain theoretical efficiency and flow capacity;
s15, optimizing the deep neural network by a Levenberg-Marquardt method to obtain an ISO standard state reduced neural network;
s16, establishing an ISO standard state conversion model of the key component through the ISO standard state conversion neural network;
s17, inputting the power plant data into an ISO standard state conversion model, and converting the efficiency and the flow capacity of key components in different running states into values under full load and ISO working conditions through the ISO standard state conversion model;
s18, transversely comparing the values of the key components under full load and ISO working conditions to obtain the efficiency and the flow capacity under the standard condition;
s2, establishing a compressor performance prediction model;
s21, establishing a deep neural network;
s22, introducing a cyclic neural network and a long-short term memory network to optimize the deep neural network to obtain a performance degradation prediction neural network;
and S23, inputting the power plant data collected in the S11 into a performance degradation prediction neural network to obtain the efficiency and the flow capacity of the compressor for predicting degradation.
2. The data-driven-based compressor performance degradation prediction algorithm of claim 1, wherein the compressor thermodynamic model comprises: compressor efficiency, compressor power consumption and compressor flow capacity;
the compressor efficiency formula is:
wherein m isaThe flow rate of the air compressor is used,is the specific enthalpy of air at the inlet temperature of the compressor,is the specific enthalpy, W, of air at the isentropic outlet temperature of the compressorc-caConsuming power for the press;
the formula of the power consumption of the compressor is as follows:
wherein m iscaThe air-extracting flow of the air compressor is the air-extracting flow,is the specific enthalpy of air at the outlet temperature of the compressor,the specific enthalpy of air at the average temperature of air extraction of the compressor is obtained;
the formula of the flow capacity of the compressor is as follows:
wherein, T1 refIs a reference value, P, of the inlet temperature of the compressor1 refIs a reference value of the compressor inlet pressure.
3. The data-driven-based compressor performance degradation prediction algorithm of claim 1, wherein: the ISO standard posture conversion model has three hidden layers, and each hidden layer comprises 256 units.
4. The data-driven-based compressor performance degradation prediction algorithm of claim 1, wherein: the power plant data comprises environmental parameters and performance parameters, the environmental parameters comprise environmental temperature, pressure and relative humidity, the performance parameters comprise power, air flow, pressure ratio, air inlet temperature and pressure of a gas turbine, inlet guide vane angle IGV and air outlet temperature and pressure of the gas turbine, and the output parameters of the ISO standard state conversion model comprise efficiency and flow capacity of the gas turbine and the gas turbine.
5. The data-driven-based compressor performance degradation prediction algorithm of claim 1, wherein optimizing the deep neural network comprises the steps of:
collecting power plant data;
preprocessing collected power plant data, wherein the preprocessing comprises the steps of filtering unsteady data of the power plant data through a sliding window method, then removing abnormal points through a Layida criterion, and finally according to the following steps of 7: and 3, randomly dividing the data into a training set and a testing set, and finally obtaining the optimized deep neural network through statistical significance difference test.
6. The data-driven-based compressor performance degradation prediction algorithm of claim 1, wherein the compressor performance prediction model has the following formula:
h(t)=φ(Ux(t)+Wh(t-1)+b)
o(t)=Vh(t)+c
wherein x is input, h is hidden unit, o is output, phi is activation function, V, W, U is weight; b. c is the deviation.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210102459.7A CN114444394A (en) | 2022-01-27 | 2022-01-27 | Gas compressor performance degradation prediction algorithm based on data driving |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210102459.7A CN114444394A (en) | 2022-01-27 | 2022-01-27 | Gas compressor performance degradation prediction algorithm based on data driving |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114444394A true CN114444394A (en) | 2022-05-06 |
Family
ID=81369454
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210102459.7A Pending CN114444394A (en) | 2022-01-27 | 2022-01-27 | Gas compressor performance degradation prediction algorithm based on data driving |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114444394A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116933693A (en) * | 2023-09-12 | 2023-10-24 | 华能南京燃机发电有限公司 | Gas turbine performance detection method and device |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102862685A (en) * | 2011-07-07 | 2013-01-09 | 中国国际航空股份有限公司 | Method for detecting whether performance of airplane components falls into degenerating stage |
US20150184549A1 (en) * | 2013-12-31 | 2015-07-02 | General Electric Company | Methods and systems for enhancing control of power plant generating units |
WO2020000248A1 (en) * | 2018-06-27 | 2020-01-02 | 大连理工大学 | Space reconstruction based method for predicting key performance parameters of transition state acceleration process of aircraft engine |
CN111754093A (en) * | 2020-06-11 | 2020-10-09 | 浙江浙能技术研究院有限公司 | Fly ash carbon content prediction method based on coal quality prediction and PSO-SVM |
-
2022
- 2022-01-27 CN CN202210102459.7A patent/CN114444394A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102862685A (en) * | 2011-07-07 | 2013-01-09 | 中国国际航空股份有限公司 | Method for detecting whether performance of airplane components falls into degenerating stage |
US20150184549A1 (en) * | 2013-12-31 | 2015-07-02 | General Electric Company | Methods and systems for enhancing control of power plant generating units |
WO2020000248A1 (en) * | 2018-06-27 | 2020-01-02 | 大连理工大学 | Space reconstruction based method for predicting key performance parameters of transition state acceleration process of aircraft engine |
CN111754093A (en) * | 2020-06-11 | 2020-10-09 | 浙江浙能技术研究院有限公司 | Fly ash carbon content prediction method based on coal quality prediction and PSO-SVM |
Non-Patent Citations (1)
Title |
---|
陈家伦;卞韶帅;黄新;: "基于BP神经网络的9F燃气轮机压气机离线水洗周期优化", 燃气轮机技术, no. 01, 16 March 2020 (2020-03-16), pages 51 - 57 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116933693A (en) * | 2023-09-12 | 2023-10-24 | 华能南京燃机发电有限公司 | Gas turbine performance detection method and device |
CN116933693B (en) * | 2023-09-12 | 2024-02-09 | 华能南京燃机发电有限公司 | Gas turbine performance detection method and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111159844B (en) | Abnormity detection method for exhaust temperature of gas turbine of power station | |
CN111648992B (en) | Gas turbine compressor fault identification early warning method | |
CN112901449B (en) | Air compressor system energy consumption optimization method based on machine learning | |
Togni et al. | A combined technique of Kalman filter, artificial neural network and fuzzy logic for gas turbines and signal fault isolation | |
CN114444394A (en) | Gas compressor performance degradation prediction algorithm based on data driving | |
CN116840720A (en) | Fuel cell remaining life prediction method | |
CN115618592A (en) | Gas path fault diagnosis method, system, equipment and terminal for gas turbine of power plant | |
CN110928248B (en) | Method for determining performance degradation degree of gas turbine | |
CN111950165B (en) | Electric energy meter reliability analysis method under typical environment | |
Ma et al. | An interpretability research of the Xgboost algorithm in remaining useful life prediction | |
CN112418306A (en) | Gas turbine compressor fault early warning method based on LSTM-SVM | |
Li et al. | Gas turbine gas‐path fault diagnosis in power plant under transient operating condition with variable geometry compressor | |
CN111310781A (en) | Refrigeration equipment refrigerant leakage identification method based on PCA-ANN | |
CN115774847A (en) | Diesel engine performance evaluation and prediction method and system | |
Losi et al. | Autoregressive Bayesian hierarchical model to predict gas turbine degradation | |
Cao et al. | Research on soft sensing modeling method of gas turbine’s difficult-to-measure parameters | |
CN117910550B (en) | Oil-free ultrahigh-speed centrifugal compressor automatic optimization system based on deep learning | |
Zhultriza et al. | Gas turbine anomaly prediction using hybrid convolutional neural network with LSTM in power plant | |
Zhang et al. | Research on gas turbine early warning based on deep learning | |
Xia et al. | Comparative Study of Two Data-Driven Gas Path Fault Quantification Methods for a Single-Shaft Gas Turbine | |
CN117807718B (en) | Aero-engine maintenance decision design method based on failure probability prediction | |
Da-li et al. | Health assessment of gas turbine performance based on SOM-MQE algorithm | |
CN118036468A (en) | Multi-mode modeling and knowledge fusion decision method for compressor washing under variable load | |
Cao et al. | Robustness optimization of gas turbine performance evaluation against sensor failures | |
Thakkar et al. | Remaining Useful Life Prediction of an Aircraft Turbofan Engine Using Deep Layer Recurrent Neural Networks. Actuators 2022, 11, 67 |
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 |