CN110849467A - Vibration monitoring method for tower type solar molten salt heat absorber - Google Patents
Vibration monitoring method for tower type solar molten salt heat absorber Download PDFInfo
- Publication number
- CN110849467A CN110849467A CN202010039663.XA CN202010039663A CN110849467A CN 110849467 A CN110849467 A CN 110849467A CN 202010039663 A CN202010039663 A CN 202010039663A CN 110849467 A CN110849467 A CN 110849467A
- Authority
- CN
- China
- Prior art keywords
- heat absorber
- freezing
- molten salt
- vector machine
- support vector
- 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
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H11/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves by detecting changes in electric or magnetic properties
- G01H11/06—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves by detecting changes in electric or magnetic properties by electric means
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Investigating Or Analyzing Materials Using Thermal Means (AREA)
Abstract
The invention relates to a vibration monitoring method for a tower type solar molten salt heat absorber, which is characterized in that vibration monitoring is carried out on a pipeline of the tower type solar molten salt heat absorber on the basis of a model of time-frequency analysis and support vector machine optimization, and a condition of frozen blockage of molten salt in the pipeline is obtained by inputting vibration signals obtained by a sensor and a signal acquisition module into the model. Early fault diagnosis and intelligent prediction of fused salt freezing blockage of the heat absorber can be realized, and the economical efficiency and safety of operation are improved; the prediction efficiency is improved, and the prediction instantaneity is ensured; the accuracy of prediction is guaranteed.
Description
Technical Field
The invention belongs to the technical field of tower-type solar molten salt heat absorbers, and particularly relates to a vibration monitoring method for a tower-type solar molten salt heat absorber.
Background
Compared with other renewable energy sources, solar energy has the advantages of unlimited reserves, universality, cleanliness in utilization, development economy and the like. The tower type solar thermal power generation system collects solar energy to a heat absorber on a high tower through a heliostat group and heats a heat transfer working medium inside the heat absorber to generate power. Compared with the solar thermal power generation technology using water as a working medium, the fused salt has the characteristics of high heat capacity, wide liquid-phase temperature range, good fluidity and the like as the heat storage working medium, and the solar thermal power generation technology using the fused salt as a heat absorption medium has the characteristics of strong heat storage capacity, stable power generation output and the like, so that the fused salt is the heat transfer and heat storage medium which is most widely applied at present in the field of photo-thermal power generation. However, the fused salt has a high freezing point, so that the risk of freezing and blocking the pipeline is possibly generated, local high temperature is caused, the heat transfer capacity of a heating surface is reduced, and the normal work of the heat absorber is damaged, so that the real-time vibration monitoring of the freezing and blocking condition of the pipeline of the heat absorber must be considered in the design and the operation, and the method has important significance for improving the safety reliability and the economical efficiency of the heat absorber.
The detection of the prior art discovers that the monitoring of the tower type molten salt heat absorber mainly focuses on the temperature monitoring of the outer screen of the heat absorber at present, more consideration is given to the heat transfer research and the structure optimization design of the heat absorber aiming at the problem of the frozen blockage of the molten salt, and almost no research is carried out on a vibration monitoring system of a pipeline of the heat absorber, so that the defects of early fault diagnosis and intelligent prediction of the heat absorber are caused, if the change trend of the fault can be predicted, the frozen blockage risk of the molten salt of the heat absorber is predicted, the maintenance or replacement of parts is more reasonably arranged, and the operation reliability and the safety of the molten salt heat absorber are greatly improved.
Disclosure of Invention
Aiming at the situation, the invention provides a vibration monitoring method for a tower-type solar molten salt heat absorber, aiming at overcoming the defects of the prior art.
In order to achieve the purpose, the invention provides the following technical scheme:
a vibration monitoring method for a tower type solar molten salt heat absorber comprises a pretreatment stage and a use stage.
The preprocessing stage comprises two parts of freezing and blocking characteristic time frequency analysis and freezing and blocking fault diagnosis of a support vector machine.
Further, the freezing and plugging characteristic time-frequency analysis is mainly divided into time domain index solution and frequency band energy distribution solution. The time domain index solving is to extract the time domain index from the time domain spectrogram of the vibration signal of the pipeline of the heat absorber under the condition of different degrees of molten salt freezing blockage; the frequency band energy distribution solving is to extract frequency band characteristics of a frequency domain spectrogram of a vibration signal of a pipeline of the heat absorber under the condition of freezing and blocking of molten salts with different degrees, and a time-frequency analysis result of the freezing and blocking characteristics is formed by combining the results of time domain index solving and frequency band energy distribution solving;
furthermore, as a plurality of frequency spectrum peak values appear in the high-frequency band of the vibration signal when the pipeline is not frozen and blocked, and after the mass of the frozen and blocked pipeline is increased, the frequency spectrum peak values are obviously weakened or even close to zero, and simultaneously the frequency spectrum peak values all appear in a relatively concentrated frequency band and do not change obviously along with the increase of the mass of the frozen and blocked pipeline, wavelet packet decomposition and extraction are carried out on the frequency band characteristics to obtain the relative energy frequency band distribution of the vibration signal;
further, the freezing and blocking fault diagnosis of the support vector machine comprises support vector machine model training and support vector machine model prediction. The obtained time-frequency analysis results of the freezing and blocking states of different pipelines are used as characteristic data input into a support vector machine model, and are firstly divided into 2 groups: training samples and prediction samples. The vector machine model training is to train the model by using the characteristics of the input training samples. And the support vector machine model prediction is to test the trained model by using a prediction sample, judge the category of the freezing and plugging fault to which the corresponding signal belongs, and compare the category with the actual category of the freezing and plugging fault of the signal to obtain the prediction accuracy of the test set. Under the condition of higher prediction accuracy of the test, the freezing and plugging condition data of the fused salt pipeline of the heat absorber is obtained through the freezing and plugging fault diagnosis of the support vector machine.
The using stage comprises three parts of signal acquisition, model based on time-frequency analysis and support vector machine optimization and information transmission.
Further, a signal acquisition part acquires vibration signals of the fused salt frozen blocked pipeline of the heat absorber through a sensor, converts the vibration signals into electric signals and inputs the electric signals into a signal acquisition module, and the signal acquisition module converts the electric signals into digital signals and stores the digital signals;
furthermore, due to the gravity effect, the molten salt is easier to freeze and block at the lower end of the pipeline, so that the sensors are arranged at the lower part of the heat absorber pipeline and are arranged in a 360-degree circular array, and one sensor is arranged on the pipeline every 60 degrees, so that the sensors can better receive vibration information;
further, predicting the signal data acquired in the use stage by using the optimized model obtained in the preprocessing stage to generate a freezing and plugging fault category result to which the signal belongs;
further, information transmission is performed by using a GPRS data transmission mechanism. The freezing and plugging fault result of the heat absorber molten salt pipeline is transmitted into the GPRS module through the serial port, and the address of the control center is obtained by using the GPRS wireless network.
The invention has the beneficial effects that:
(1) according to the method, vibration monitoring is carried out on the pipeline of the tower type solar molten salt heat absorber based on a model of time-frequency analysis and support vector machine optimization, and the condition of frozen blockage of the molten salt in the pipeline is obtained by inputting vibration signals obtained by a sensor and a signal acquisition module into the model. Compared with the method for avoiding the frozen blockage fault from the perspective of heat transfer research and structural optimization design of the heat absorber, the method is simple to operate, does not need to relate to complex design, can realize early fault diagnosis and intelligent prediction of the fused salt frozen blockage of the heat absorber, and improves the economical efficiency and safety of operation;
(2) the method is divided into a preprocessing stage and a using stage, wherein the preprocessing stage is used for training a model based on time-frequency analysis and support vector machine optimization by using different molten salt freezing and plugging signal data obtained by experimental measurement, the preprocessing stage does not consume actual time, and only the trained model is used in actual application, so that the prediction efficiency is improved, and the prediction instantaneity is ensured;
(3) the method uses a model based on time-frequency analysis and support vector machine optimization to extract the characteristics of the vibration signal of the pipeline, and then outputs a prediction result with higher accuracy through training and prediction of the support vector machine model, thereby ensuring the accuracy of prediction.
Drawings
FIG. 1 is a pretreatment stage of the vibration monitoring method for the tower-type solar molten salt heat absorber.
FIG. 2 is a use stage of the vibration monitoring method for the tower-type solar molten salt heat absorber.
Fig. 3 is a time-frequency diagram of different freezing and plugging states of a heat absorber pipeline obtained according to a set of experimental data in the embodiment.
FIG. 4 is a graph showing the prediction accuracy of the trained support vector machine model in the example.
Fig. 5 is a sensor layout diagram of the vibration monitoring method for the tower-type solar molten salt heat absorber.
Detailed Description
The technical solutions of the present invention are described in detail below with reference to the accompanying drawings and specific embodiments, and it should be noted that the detailed description is only for describing the present invention and should not be construed as limiting the present invention.
The invention discloses a vibration monitoring method for a tower type solar molten salt heat absorber.
As shown in fig. 1, the pre-treatment stage includes the following two parts:
and (3) freeze plugging characteristic time-frequency analysis: respectively carrying out time domain index solution and frequency band energy distribution solution on the obtained pipeline vibration signals, and combining the two solution results to form a time-frequency analysis result of the freezing and plugging characteristic, wherein the time-frequency analysis result is used as characteristic data for the freezing and plugging fault diagnosis of the support vector machine at the next stage;
further, the time domain index solving is to extract the time domain index of the time domain spectrogram of the vibration signal of the pipeline of the heat absorber under the condition of freezing and blocking of the molten salt with different degrees, including peak valueX p Mean value ofRoot mean square valueX rms The calculation formulas are respectively shown as formulas (1) - (3):
(2)
in the above formula, the vibration signal is set asX i (i = 1-n, n being the number of nodes);
the frequency band energy distribution solving is to extract frequency band characteristics of a frequency domain spectrogram of a vibration signal of a pipeline of the heat absorber under the condition of freezing and blocking of molten salts with different degrees, and the calculation mode is as follows:
the spectral energy can be calculated by equation (4):
in the formula (I), the compound is shown in the specification,j=0,1…,2 i -1,k=1,2,…,N,Nin order to be the number of the nodes,the vibration signal is decomposed by a wavelet packetiLayer onejBand energy of individual nodes;
the total frequency band energy of the vibration signal can be calculated by equation (5):
the relative band energy of the frequency band corresponding to each node can be calculated by equation (6):
and (3) diagnosing the freezing and blocking faults of the support vector machine: and dividing the time frequency result characteristic data obtained in the last part into a training sample and a prediction sample. Firstly, inputting a training sample, and training a model of a support vector machine to obtain a trained model of the support vector machine; and secondly, inputting a test sample into the trained model of the support vector machine for testing, judging the category of the freezing and plugging fault to which the corresponding signal belongs, and comparing the result with the actual category of the freezing and plugging fault of the signal to obtain the prediction accuracy of the test set. And when the prediction accuracy reaches a higher level, completing the model preprocessing stage based on time-frequency analysis and support vector machine optimization.
As shown in fig. 2, the use phase includes the following three steps:
signal acquisition: the heat absorber molten salt pipeline generates vibration, the sensor senses the vibration and inputs an electric signal to the signal acquisition module, and the signal acquisition module converts an analog signal into a digital signal and stores the digital signal;
model based on time-frequency analysis and support vector machine optimization: substituting the acquired signal data into the model to obtain frozen plugging category data of the heat absorber molten salt pipeline, and realizing monitoring and prediction of frozen plugging faults;
as a preferred mode, a group of experimental data is selected and substituted for preprocessing, a frequency domain diagram of different freezing and plugging states of the heat absorber pipeline can be obtained as shown in fig. 3, then the experimental data is divided into training samples and testing samples, the training samples are substituted into a model for training, and the prediction accuracy generated by comparing the training samples with the testing samples is shown in fig. 4, so that a better prediction effect can be seen, and the final training accuracy can reach 0.997. Information transmission: using the GPRS data transmission mechanism. And the freezing and blocking fault result of the fused salt pipeline of the heat absorber is transmitted into the GPRS module through the serial port, and the address of the control center is obtained by utilizing the GPRS wireless network.
Further, as shown in fig. 5, the sensors in the signal acquisition step are arranged at the lower end of the pipeline of the heat absorber in a 360-degree circular array, and one sensor is arranged on the pipeline every 60 degrees, so that due to the action of gravity, the frozen plugging of the molten salt is easy to occur at the lower end of the pipeline, and the sensors are arranged at the lower end of the pipeline to better receive vibration information.
Claims (3)
1. A vibration monitoring method for a tower-type solar molten salt heat absorber is characterized by comprising a pretreatment stage and a use stage;
the preprocessing stage comprises two parts of freezing and plugging characteristic time frequency analysis and freezing and plugging fault diagnosis of a support vector machine;
and (3) freeze plugging characteristic time-frequency analysis: respectively carrying out time domain index solution and frequency band energy distribution solution on the obtained pipeline vibration signals, and combining the two solution results to form a time-frequency analysis result of the freezing and plugging characteristic, wherein the time-frequency analysis result is used as characteristic data for the freezing and plugging fault diagnosis of the support vector machine at the next stage;
and (3) diagnosing the freezing and blocking faults of the support vector machine: dividing the time frequency result characteristic data obtained in the previous part into a training sample and a prediction sample; firstly, inputting a training sample, and training a model of a support vector machine to obtain a trained model of the support vector machine; secondly, inputting a test sample into a trained model of the support vector machine for testing, judging the category of the freezing and plugging fault to which the corresponding signal belongs, and comparing the result with the actual category of the freezing and plugging fault of the signal to obtain the prediction accuracy of a test set; when the prediction accuracy reaches a higher level, completing a model preprocessing stage based on time-frequency analysis and support vector machine optimization;
the using stage comprises three steps of signal acquisition, model based on time-frequency analysis and support vector machine optimization and information transmission;
signal acquisition: the heat absorber molten salt pipeline generates vibration, the sensor senses the vibration and inputs an electric signal to the signal acquisition module, and the signal acquisition module processes the electric signal to realize display and storage of the electric signal;
model based on time-frequency analysis and support vector machine optimization: and substituting the acquired signal data into the model to obtain the frozen plugging category data of the heat absorber molten salt pipeline, thereby realizing the monitoring and prediction of frozen plugging faults.
2. The vibration monitoring method for the tower-type solar molten salt heat absorber according to claim 1, wherein the freezing and blocking characteristic time-frequency analysis in the preprocessing stage comprises time domain index solution and frequency band energy distribution solution; the time domain index solving is to extract the time domain index from the time domain spectrogram of the vibration signal of the pipeline of the heat absorber under the condition of different degrees of molten salt freezing blockage; the frequency band energy distribution solving is to extract frequency band characteristics of a frequency domain spectrogram of a vibration signal of a pipeline of the heat absorber under the condition of freezing and blocking of molten salts with different degrees.
3. The vibration monitoring method for the tower-type solar molten salt heat absorber according to claim 1, characterized in that the information transmission: the GPRS data transmission mechanism is utilized for carrying out the operation; and the freezing and blocking fault result of the fused salt pipeline of the heat absorber is transmitted into the GPRS module through the serial port, and the address of the control center is obtained by utilizing the GPRS wireless network.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010039663.XA CN110849467A (en) | 2020-01-15 | 2020-01-15 | Vibration monitoring method for tower type solar molten salt heat absorber |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010039663.XA CN110849467A (en) | 2020-01-15 | 2020-01-15 | Vibration monitoring method for tower type solar molten salt heat absorber |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110849467A true CN110849467A (en) | 2020-02-28 |
Family
ID=69610754
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010039663.XA Pending CN110849467A (en) | 2020-01-15 | 2020-01-15 | Vibration monitoring method for tower type solar molten salt heat absorber |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110849467A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111307493A (en) * | 2020-05-11 | 2020-06-19 | 杭州锅炉集团股份有限公司 | Knowledge-based fault diagnosis method for tower type solar molten salt heat storage system |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1584519A (en) * | 2004-06-15 | 2005-02-23 | 东北师范大学 | Dynamic measuring system with four sensors for acoustic field spacial distribution of acho signals of bat |
CN101718581A (en) * | 2009-11-13 | 2010-06-02 | 浙江大学 | Alarming method of nuclear power station loose part based on support vector machine |
CN103256974A (en) * | 2013-04-15 | 2013-08-21 | 北京天诚同创电气有限公司 | Internally-arranged FFT on-line frequency detection module application |
CN103335840A (en) * | 2013-07-02 | 2013-10-02 | 中煤科工集团西安研究院 | Intelligent diagnosis method for faults of mining drilling machine gearbox |
-
2020
- 2020-01-15 CN CN202010039663.XA patent/CN110849467A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1584519A (en) * | 2004-06-15 | 2005-02-23 | 东北师范大学 | Dynamic measuring system with four sensors for acoustic field spacial distribution of acho signals of bat |
CN101718581A (en) * | 2009-11-13 | 2010-06-02 | 浙江大学 | Alarming method of nuclear power station loose part based on support vector machine |
CN103256974A (en) * | 2013-04-15 | 2013-08-21 | 北京天诚同创电气有限公司 | Internally-arranged FFT on-line frequency detection module application |
CN103335840A (en) * | 2013-07-02 | 2013-10-02 | 中煤科工集团西安研究院 | Intelligent diagnosis method for faults of mining drilling machine gearbox |
Non-Patent Citations (1)
Title |
---|
彭梁: "基于炉外管屏振动信号分析的燃煤锅炉过热器结渣诊断研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111307493A (en) * | 2020-05-11 | 2020-06-19 | 杭州锅炉集团股份有限公司 | Knowledge-based fault diagnosis method for tower type solar molten salt heat storage system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104929864A (en) | Field programmable gate array (FPGA)-based embedded type operating state monitoring and fault diagnosis system for wind generating set | |
CN108983749B (en) | Photovoltaic array fault diagnosis method based on K-SVD training sparse dictionary | |
CN105137242A (en) | Single-phase photovoltaic inverter on-line state monitoring and residual life prediction method | |
CN109993445A (en) | A kind of integrated energy system vulnerability assessment method considering photovoltaic prediction error | |
CN104749999A (en) | Accurate guidance system for optimizing operation of turbo generator group cold end system of assembly wet-type cooling tower | |
CN102494895B (en) | Analyzing method for energy saving and optimization of steam turbine set of power station | |
CN105244890A (en) | Reactive power optimization method for new energy grid connection | |
CN110849467A (en) | Vibration monitoring method for tower type solar molten salt heat absorber | |
CN104574221B (en) | A kind of photovoltaic plant running status discrimination method based on loss electricity characteristic parameter | |
CN113297742B (en) | Self-energy non-invasive system based on independent component analysis and modeling method thereof | |
CN202903711U (en) | Heat loss detection device of high-temperature vacuum solar heat-collecting tube | |
CN104617578A (en) | Method for acquiring available power transmission capability of power system with wind power plant | |
Yang | Towards the development of a wake meandering model based on neural networks | |
Pouraltafi-kheljan et al. | Power generation nowcasting of the behind-the-meter photovoltaic systems | |
CN105528517B (en) | Based on neural network, wavelet decomposition predicting power of photovoltaic plant method, system | |
Han et al. | Electrical performance and power prediction of a roll-bond photovoltaic thermal array under dewing and frosting conditions | |
Guo et al. | Early fault detection of wind turbine gearbox based on Adam-trained LSTM | |
CN112907028B (en) | Heat exchange system transient working condition energy consumption analysis method based on energy potential of media | |
CN110957723B (en) | Data-driven method for rapidly evaluating transient voltage safety of power grid on line | |
CN111307493B (en) | Knowledge-based fault diagnosis method for tower type solar molten salt heat storage system | |
Xie et al. | Performance optimization of the air-cooling system in a coal-fired power unit based on intelligent algorithms | |
CN105353820A (en) | Control method for tracking MPPT device by maximum powers point used for photovoltaic cell array | |
Yu et al. | Experimental study on performance of trough solar thermal power station | |
Ma et al. | Fault Diagnosis of Integrated Energy System Based on CNN-LSTM | |
Akbar et al. | Solar Thermal Process Parameters Forecasting for Evacuated Tubes Collector (ETC) Based on RNN-LSTM |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200228 |