CN111897203A - Heater economic regulation method based on data driving and mechanism modeling state monitoring - Google Patents
Heater economic regulation method based on data driving and mechanism modeling state monitoring Download PDFInfo
- Publication number
- CN111897203A CN111897203A CN202010554012.4A CN202010554012A CN111897203A CN 111897203 A CN111897203 A CN 111897203A CN 202010554012 A CN202010554012 A CN 202010554012A CN 111897203 A CN111897203 A CN 111897203A
- Authority
- CN
- China
- Prior art keywords
- data
- heater
- model
- end difference
- economic
- 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
- 238000000034 method Methods 0.000 title claims abstract description 47
- 230000007246 mechanism Effects 0.000 title claims abstract description 29
- 238000012544 monitoring process Methods 0.000 title claims abstract description 24
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 21
- 238000004364 calculation method Methods 0.000 claims abstract description 19
- 238000004458 analytical method Methods 0.000 claims abstract description 15
- 238000000605 extraction Methods 0.000 claims abstract description 3
- 238000007689 inspection Methods 0.000 claims abstract 2
- 238000012546 transfer Methods 0.000 claims description 9
- 238000012549 training Methods 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 6
- 238000000513 principal component analysis Methods 0.000 claims description 5
- 230000009467 reduction Effects 0.000 claims description 3
- 238000010946 mechanistic model Methods 0.000 claims description 2
- 230000003750 conditioning effect Effects 0.000 claims 1
- 230000008569 process Effects 0.000 abstract description 10
- 230000008859 change Effects 0.000 abstract description 9
- 230000008878 coupling Effects 0.000 abstract description 7
- 238000010168 coupling process Methods 0.000 abstract description 7
- 238000005859 coupling reaction Methods 0.000 abstract description 7
- 230000001502 supplementing effect Effects 0.000 abstract description 3
- 238000011217 control strategy Methods 0.000 abstract description 2
- 230000001172 regenerating effect Effects 0.000 abstract 1
- 239000013589 supplement Substances 0.000 abstract 1
- 238000010998 test method Methods 0.000 description 4
- 238000003745 diagnosis Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000000844 transformation Methods 0.000 description 1
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
- G05B11/00—Automatic controllers
- G05B11/01—Automatic controllers electric
- G05B11/36—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential
- G05B11/42—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential for obtaining a characteristic which is both proportional and time-dependent, e.g. P.I., P.I.D.
Abstract
The invention provides a heater economic regulation method based on data driving and mechanism modeling state monitoring, which comprises the steps of firstly carrying out steady state judgment on regenerative system operation data containing a plurality of variables through PCA and R inspection methods, extracting characteristic indexes, obtaining end difference operation steady state values under different water supply temperatures and steam extraction pressures, then establishing a model capable of dynamically identifying the lower end difference over-limit value during load change based on a deep stacking self-coding method, obtaining the end difference and water level characteristic relation by combining a numerical calculation model established by heater mechanism analysis, obtaining a water level deviation value from the end difference over-limit value, and finally adopting a PID control strategy to regulate the water level deviation to realize the economic operation of a unit. The invention realizes the combination of data modeling and mechanism modeling, can establish a monitoring model meeting the characteristics of the operation process, can embody the grasp of the mechanism model on the variable coupling relation, and supplements the accuracy of the model prediction by a data supplementing method under the condition of no historical data.
Description
Technical Field
The invention relates to the field of artificial intelligence data-driven fault diagnosis, and provides a heater economic regulation method based on data driving and mechanism modeling state monitoring by combining mechanism analysis numerical computation modeling.
Background
Considering that most modern industrial processes have more than three process variables, complex model structures and coupling relations, the relation among the variables is nonlinear, various sampling frequencies can exist, and the like, so that an accurate physical model cannot be established for analysis by a traditional method. The data-driven performance diagnosis method based on artificial intelligence can detect abnormal conditions occurring in the industrial process in real time and can acquire system characteristic parameters with associated characteristics from historical data. Therefore, the modeling and diagnosis requirements of the thermal process system can be met under the condition that the complex mechanism characteristics of the system are not deeply known.
However, considering that data-driven modeling analysis needs to be established on the basis of existing data, and existing data are distributed in a certain interval, the accuracy can be ensured by using a model trained by historical data in the interval. The prediction of the system change trend of the model after the exceeding interval does not necessarily accord with the reality. Although the establishment process of the mechanism model is complex, the model grasps the coupling relation among all variables and is embodied by a certain formula, so that the change trend of the system can be accurately predicted under the condition of lacking of actual operation data. Therefore, a mechanism model is established, and the further change trend of the system can be researched at a place beyond the existing data interval. The mechanism model is combined with the data model, so that the research interval can be widened, and the characteristic change of the system can be more accurately predicted.
Disclosure of Invention
According to the invention, through analysis and processing of historical data of the power plant, the lower end difference of the heater is taken as an entry point, an economic and safety monitoring model of the heater is established by a data driving method based on the stacking automatic encoder, and a deviation value of the lower end difference from the optimal operation condition under different loads can be dynamically identified. And then, combining the end difference and water level characteristic relation obtained by a numerical calculation model established by mechanism analysis, and obtaining a water level change value from the end difference deviation value. The combination of data modeling and mechanism modeling is realized, a data driving method can be used for analyzing a large amount of historical data to establish a monitoring model meeting the characteristics of an operation process, the grasp of a mechanism model on a variable coupling relation can be embodied, and the accuracy of model prediction is realized by a data supplementing method under the condition of no historical data.
In order to achieve the above purpose, the invention provides the following technical scheme:
the economic regulation method of the heater based on data driving and mechanism modeling state monitoring is characterized in that: and performing dimensionality reduction on historical data of the power plant by using a PCA analysis method, and judging the steady state of the data by combining an R test method to obtain end difference operation steady state values under different water supply temperatures and steam extraction pressures. And then, based on a Python writing program, training the stacked self-encoder by using steady-state data in a greedy layer-by-layer pre-training mode, and establishing a dynamic monitoring model of endplay. The mechanism model is established on the basis of three-stage heat transfer of the heater, and a numerical calculation model is obtained by using MATLAB programming according to a heat transfer empirical formula of literature data. Substituting data under different working conditions for calculation to obtain a large number of system results of numerical calculation, and establishing a characteristic relation between the end difference and the water level. And finally, on the basis of characteristic analysis of the high-pressure heater, obtaining the optimal PID control adjusting parameter through a step disturbance test, and adjusting the water level deviation to realize economic operation of the heater.
Furthermore, historical data processing is based on an original PCA analysis method, and whether the multivariable system is in a stable state or not is judged by combining an R test method.
Furthermore, considering that the self-coding network has a certain deviation in learning the characteristic indexes, the stacked self-coder has more excellent characteristic learning capability.
Furthermore, each section of the three-section heat transfer calculation model of the heater selects 1000 discrete nodes, the requirement of model precision can be met, and the calculation amount is not too large.
Further, in the adjusting process of the high water adding level, the optimal PID control adjusting parameter is obtained through a large number of step disturbance tests.
Compared with the prior art, the invention has the beneficial effects that:
the state monitoring model established by using the data method is well matched with the dynamic characteristics of the system only in a certain interval, and the change trend of the system is not necessarily accurate in the interval without training data; the method is combined with data and a mechanism method to establish a model, a data method is used for processing historical operation data to establish a heater end difference dynamic monitoring model, a numerical value calculation method is used for establishing a mechanism model, the coupling relation among all variables is grasped, and the monitoring model of the data method is combined to carry out comprehensive analysis to obtain the change value of the water level of the heater, so that the economic adjustment of the heater is realized.
Drawings
FIG. 1 is a flow chart of a heater economy tuning method based on data-driven and mechanistic modeling state monitoring provided by the present invention.
FIG. 2 is a flow chart for processing historical plant operating data using PCA and R test methods.
FIG. 3 is a flow chart for building a dynamic monitoring model using a stacked self-encoding method based on historical data.
FIG. 4 is a flow chart of a numerical calculation model established using a mechanism analysis method based on a three-stage heat transfer model.
Detailed Description
The invention is further described below with reference to the accompanying drawings.
As shown in fig. 1, historical data is processed for a data driving method in a left virtual frame, and a dynamic end-to-end difference monitoring model process is established by using Python programming; analyzing heat transfer modes of all parts of the heater by a mechanism method in a middle virtual frame, grasping coupling relations among all variables, and establishing a numerical calculation model through discrete node iterative calculation to obtain a terminal difference water level characteristic relation; and the right virtual frame is used for analyzing the characteristics of the high-pressure heater and combining a step disturbance test to obtain an optimal PID control regulation parameter so as to realize the economic regulation of the heater.
The invention provides a high-pressure heater economic regulation method based on data driving and mechanism modeling state monitoring. And then training the stacked self-encoder by using steady-state data to establish a model capable of dynamically identifying the lower end difference exceeding the limit value when the load changes. And obtaining a water level deviation value from the end difference over-limit value by combining the end difference and water level characteristic relation obtained by a numerical calculation model established by heater mechanism analysis. And finally, regulating the water level deviation by adopting a PID control strategy to realize the economic operation of the unit.
Fig. 2 shows a method for processing history data. Firstly, data preprocessing is carried out, and historical data { x }iAnd normalizing to eliminate the influence of the index dimension. Then calling PCA to normalize the variable
Using PCA, the transformed variables { y are obtainedi}. Then selecting dimension reduction data according to variable { yiThe weight of each dimension, the k-dimensional pivot of the analysis is preserved. And finally, judging by using an R test method, and performing steady-state analysis on the weighted variables. By processing the historical data, the operation data of the unit under the optimal stable working condition can be obtained and used as training data to establish a model.
FIG. 3 shows a process for building a dynamic monitoring model using a stacked self-encoding approach based on steady-state historical data. And training the stacked self-encoder to learn the optimal value of the end difference according to the steady-state data, establishing a dynamic monitoring model of the end difference under different loads, and using Python programming to realize the end difference. The stacked self-encoder is trained and then evaluated using Root Mean Square Error (RMSE). RMSE is a parameter representing the difference between input data and output (predicted) data obtained by the model. As shown in formula (1.1).
In the formula: n is the number of groups of data, N is the dimension of each group of data, xpkAnd x'pkRespectively, input data and prediction data output after being stacked from the encoder.
Fig. 4 shows a flow of establishing a heater numerical calculation model by selecting discrete nodes for iterative calculation based on a three-stage heat transfer model through mechanism analysis. The numerical calculation model is substituted into the first high-voltage design data of the 600MW unit for simulation, and on the basis of further grasping the characteristics of the heater, the corresponding relation between the performance index (water level) and the quality index (lower end difference) is established. The defect that the data method established model cannot predict the system change trend is overcome.
The control rule of the high water adding position PID control system is as the following formula (1.2):
in the formula: kpIs a proportional gain factor, TiIs the integration time constant, TdIs a differential time constant, u (t) is a control quantity, e (t) is a difference between a set value and an actual value. In the actual operation process, control simulation is carried out through corresponding curves of steps, and an adjusting parameter K is determinedp、TiAnd Td. The system can be adjusted, and economical adjustment of the water level is achieved.
In summary, by applying the economic regulation method of the high-pressure heater based on data driving and mechanism modeling state monitoring, provided by the invention, a large amount of historical data can be analyzed by using the data driving method, a monitoring model meeting the characteristics of an operation process is established, the grasp of a mechanism model on a variable coupling relation can be reflected, and the accuracy of model prediction is realized by a data supplementing method under the condition that no historical data exists.
The method and the specific implementation method of the invention are described in detail above, and the corresponding implementation flow is given. Of course, the present invention may have other embodiments besides the above-mentioned examples, and all the technical solutions formed by using equivalent substitutions or equivalent transformations fall within the protection scope of the present invention.
Claims (4)
1. The economic regulation method of the heater based on data driving and mechanism modeling state monitoring is characterized by comprising the following steps: performing dimensionality reduction on historical data of the power plant by using a PCA (principal component analysis) analysis method, and judging the steady state of the data by combining an R (R) inspection method to obtain end difference operation steady state values under different water supply temperatures and steam extraction pressures; then, based on a Python writing program, training a stacked self-encoder by using steady-state data in a greedy layer-by-layer pre-training mode, and establishing a dynamic monitoring model of endplay; the mechanism model is established on the basis of three-section heat transfer of the heater, and a numerical calculation model is obtained by using MATLAB programming according to a heat transfer empirical formula; substituting data under different working conditions for calculation to obtain a large number of system results of numerical calculation, and establishing a characteristic relation between the end difference and the water level; and finally, on the basis of characteristic analysis of the high-pressure heater, obtaining the optimal PID control adjusting parameter through a step disturbance test, and adjusting the water level deviation to realize economic operation of the heater.
2. The method of claim 1 wherein the historical data processing is based on PCA analysis and incorporates R-test to determine if the multivariate system is in steady state.
3. The economic conditioning method of a heater based on data-driven and mechanistic modeling state monitoring of claim 1, characterized in that 1000 discrete sections are selected for each section of the three-section heat transfer calculation model of the heater.
4. The economic regulation method of a heater based on data driving and mechanism modeling state monitoring as claimed in claim 1, characterized in that during the adjustment of high water-adding level, the optimal PID control regulation parameter is obtained through a large number of step disturbance tests.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010554012.4A CN111897203A (en) | 2020-06-17 | 2020-06-17 | Heater economic regulation method based on data driving and mechanism modeling state monitoring |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010554012.4A CN111897203A (en) | 2020-06-17 | 2020-06-17 | Heater economic regulation method based on data driving and mechanism modeling state monitoring |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111897203A true CN111897203A (en) | 2020-11-06 |
Family
ID=73206758
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010554012.4A Pending CN111897203A (en) | 2020-06-17 | 2020-06-17 | Heater economic regulation method based on data driving and mechanism modeling state monitoring |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111897203A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112503789A (en) * | 2020-12-09 | 2021-03-16 | 西安交通大学 | Intermediate pressure control method for cascade refrigeration system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106765047A (en) * | 2017-01-11 | 2017-05-31 | 中国神华能源股份有限公司 | Water level regulation method and apparatus |
| CN108494021A (en) * | 2018-04-20 | 2018-09-04 | 东北大学 | The Stability Assessment of electric-thermal-gas integrated energy system and static cost control method |
| CN110008575A (en) * | 2019-03-29 | 2019-07-12 | 重庆大学 | Recirculating cooling water system processing medium multi-temperature target set point switches multi-parameter prediction control algolithm |
| CN111077869A (en) * | 2019-12-31 | 2020-04-28 | 国电九江发电有限公司 | Optimization control method and system for big data intelligent control bag-type dust collector |
-
2020
- 2020-06-17 CN CN202010554012.4A patent/CN111897203A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106765047A (en) * | 2017-01-11 | 2017-05-31 | 中国神华能源股份有限公司 | Water level regulation method and apparatus |
| CN108494021A (en) * | 2018-04-20 | 2018-09-04 | 东北大学 | The Stability Assessment of electric-thermal-gas integrated energy system and static cost control method |
| CN110008575A (en) * | 2019-03-29 | 2019-07-12 | 重庆大学 | Recirculating cooling water system processing medium multi-temperature target set point switches multi-parameter prediction control algolithm |
| CN111077869A (en) * | 2019-12-31 | 2020-04-28 | 国电九江发电有限公司 | Optimization control method and system for big data intelligent control bag-type dust collector |
Non-Patent Citations (2)
| Title |
|---|
| 李萍: "大型凝汽式汽轮机系统设备建模与运行初压优化", 《中国优秀硕士学位论文全文数据库工程科技II辑》 * |
| 雷丽君: "大型火电机组高压给水加热器建模与故障诊断", 《中国优秀硕士学位论文全文数据库工程科技II辑》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112503789A (en) * | 2020-12-09 | 2021-03-16 | 西安交通大学 | Intermediate pressure control method for cascade refrigeration system |
| CN112503789B (en) * | 2020-12-09 | 2021-09-03 | 西安交通大学 | Intermediate pressure control method for cascade refrigeration system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104537415B (en) | A kind of non-linear process industry failure predication and recognition methodss based on compressed sensing and DROS ELM | |
| CN110530650B (en) | Method for monitoring performance state of heavy-duty gas turbine based on generalized regression neural network and box diagram analysis | |
| CN112101480A (en) | Multivariate clustering and fused time sequence combined prediction method | |
| CN109670625B (en) | NOx emission concentration prediction method based on unscented Kalman filtering least square support vector machine | |
| CN105955026B (en) | Fuzzy PID control method, device and system | |
| CN108121215B (en) | Process control loops method of evaluating performance and device based on full loop reconstruct emulation | |
| CN109032117B (en) | ARMA model-based single-loop control system performance evaluation method | |
| CN110413601B (en) | Generator data screening method based on Gauss naive Bayes and prediction error method | |
| CN105912822B (en) | The neuron network simulation method of seawater desulfurization device desulfuration efficiency | |
| CN114117919B (en) | Instant learning soft measurement modeling method based on sample collaborative representation | |
| CN115688581A (en) | Oil gas gathering and transportation station equipment parameter early warning method, system, electronic equipment and medium | |
| CN111897203A (en) | Heater economic regulation method based on data driving and mechanism modeling state monitoring | |
| CN117031950A (en) | Modeling method and device for deep peak-shaving thermal power generating unit control system | |
| CN116430726A (en) | Method for controlling turboset based on subtractive clustering and fuzzy neural network | |
| Yang et al. | Application of fuzzy neural network PID algorithm in oil pump control | |
| CN116662925A (en) | Industrial process soft measurement method based on weighted sparse neural network | |
| CN116484675A (en) | Crack propagation life prediction method and system for ship engine blade | |
| CN111142488A (en) | Process industrial optimization method based on working condition comparison driving | |
| Yordanova | Intelligent approaches to real time level control | |
| CN114692729A (en) | New energy station bad data identification and correction method based on deep learning | |
| CN113088359A (en) | Triethylene glycol loss online prediction method of triethylene glycol dehydration device driven by technological parameters | |
| CN112748663A (en) | Wind power torque fault-tolerant control method based on data-driven output feedback | |
| CN111695300A (en) | Steam turbine valve flow fitting method and system | |
| CN114326395B (en) | Intelligent generator set control model online updating method based on working condition discrimination | |
| CN113806996B (en) | Grate cooler optimization method based on time memory modeling |
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 |
Application publication date: 20201106 |
|
| RJ01 | Rejection of invention patent application after publication |