CN112906307A - Steam yield prediction method of steam boiler based on data mining - Google Patents
Steam yield prediction method of steam boiler based on data mining Download PDFInfo
- Publication number
- CN112906307A CN112906307A CN202110317633.5A CN202110317633A CN112906307A CN 112906307 A CN112906307 A CN 112906307A CN 202110317633 A CN202110317633 A CN 202110317633A CN 112906307 A CN112906307 A CN 112906307A
- Authority
- CN
- China
- Prior art keywords
- data
- steam
- model
- training
- boiler
- 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 35
- 238000007418 data mining Methods 0.000 title claims abstract description 17
- 238000012549 training Methods 0.000 claims abstract description 49
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 22
- 238000004519 manufacturing process Methods 0.000 claims abstract description 22
- 230000004927 fusion Effects 0.000 claims abstract description 17
- 238000012360 testing method Methods 0.000 claims abstract description 12
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 22
- 238000007637 random forest analysis Methods 0.000 claims description 15
- 239000007789 gas Substances 0.000 claims description 14
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 claims description 14
- 239000003345 natural gas Substances 0.000 claims description 7
- 238000012417 linear regression Methods 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 5
- 238000012795 verification Methods 0.000 claims description 5
- 230000008020 evaporation Effects 0.000 claims description 4
- 238000001704 evaporation Methods 0.000 claims description 4
- 238000004140 cleaning Methods 0.000 claims description 3
- 238000003066 decision tree Methods 0.000 claims description 3
- 238000009826 distribution Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 235000010724 Wisteria floribunda Nutrition 0.000 claims description 2
- 238000013501 data transformation Methods 0.000 claims description 2
- 230000010354 integration Effects 0.000 claims description 2
- 230000000717 retained effect Effects 0.000 claims description 2
- 238000009827 uniform distribution Methods 0.000 claims description 2
- 230000008569 process Effects 0.000 abstract description 4
- 238000011156 evaluation Methods 0.000 abstract 1
- 230000003044 adaptive effect Effects 0.000 description 7
- 238000013528 artificial neural network Methods 0.000 description 7
- 238000002485 combustion reaction Methods 0.000 description 6
- 238000012706 support-vector machine Methods 0.000 description 6
- 239000000446 fuel Substances 0.000 description 4
- 239000000779 smoke Substances 0.000 description 4
- 230000006872 improvement Effects 0.000 description 2
- 238000007500 overflow downdraw method Methods 0.000 description 2
- 241000208125 Nicotiana Species 0.000 description 1
- 235000002637 Nicotiana tabacum Nutrition 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 239000003344 environmental pollutant Substances 0.000 description 1
- 239000002737 fuel gas Substances 0.000 description 1
- 239000000295 fuel oil Substances 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000003921 oil Substances 0.000 description 1
- 210000004205 output neuron Anatomy 0.000 description 1
- 231100000719 pollutant Toxicity 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 239000002918 waste heat Substances 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
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/22—Yield analysis or yield optimisation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Feedback Control In General (AREA)
Abstract
The invention discloses a data mining-based steam yield prediction method for a steam boiler, and relates to the technical field of steam yield production of a gas steam boiler. The method analyzes the steam generation process, and establishes a steam yield prediction model by adopting a data mining algorithm on the basis of the sensor data of the steam boiler, wherein the steam yield prediction model can well predict the steam yield. The invention selects model parameters by using a boosting-tree algorithm, establishes a steam yield prediction model by adopting a self-adaptive weight fusion model, and comprehensively tests and evaluates the performance of the prediction model on the basis of 5 evaluation indexes. The input data is simplified, the data training time is saved, and the prediction precision is higher.
Description
Technical Field
The invention relates to the technical field of steam yield production of gas steam boilers, in particular to a method for predicting steam yield of a steam boiler based on data mining.
Background
The gas-fired steam boiler refers to a steam boiler heated by gas combustion. The vertical steam boiler adopts a mode of arranging the burner at the bottom and has a two-return structure, fuel is fully combusted, the boiler is stable in operation and small in occupied space, and meanwhile, the spoilers are inserted into the smoke pipes, so that the smoke exhaust speed is reduced, the heat exchange quantity is increased, the thermal efficiency of the boiler is high, and the use cost of a user is reduced; horizontal steam boiler is the full wet back of a boiler following current three return stroke smoke and fire tube structure of shell type, and flame is the pressure fired in big combustion chamber pressure-fired a little, and the complete extension, the combustion heat load is low, and the combustion heat efficiency is high, has reduced exhaust gas temperature effectively, and energy saving and consumption reduction uses more economically, adopts wave form furnace pipe and screw thread tobacco pipe structure, has improved the heat absorption intensity of boiler promptly, has satisfied the heat transfer surface and has heated the expanded demand again, scientific and reasonable, durable.
The energy-saving mode of the gas steam boiler is more, and the traditional energy-saving mode focuses more on the performance improvement of the steam boiler. For example: the rated power and the number of the boilers of the natural gas boiler are reasonably selected according to the amount of steam required by industrial production, and the higher the matching degree of the two conditions with actual use is, the smaller the smoke loss is, and the more obvious the energy is saved; the fuel is fully contacted with the air, and the proper amount of fuel and the proper amount of air form the optimal proportion for combustion, so that the combustion efficiency of the fuel can be improved, the emission of pollutant gas can be reduced, and the aim of dual energy conservation can be realized; the exhaust gas temperature of the natural gas boiler is reduced, the exhaust gas temperature of the boiler is reduced, waste heat generated in the exhaust gas is effectively utilized, the efficiency of the common boiler is 85-88%, and the exhaust gas temperature is 220-. If an economizer is arranged and the exhaust gas temperature is reduced to 150 ℃ after the heat of the exhaust gas is utilized, the boiler efficiency can be improved to about 90-93 percent, and other energy-saving measures are supplemented, so that the boiler thermal efficiency can easily reach more than 95 percent; the heat of the boiler blow-off water is recycled, and the heat of the continuous blow-off water is utilized through heat exchange, so that the water supply temperature of the deoxygenated water is increased, and the purpose of saving energy of the natural gas steam boiler is achieved; and condensed water in the heat supply system can be reasonably recovered, and the heat of the condensed water can be recycled.
However, the above energy-saving methods can quickly reach the energy-saving bottleneck because the methods are influenced by the process of the steam boiler. Meanwhile, when the load of the boiler varies, the boiler may not be maintained in an operating state in which the thermal efficiency is the highest. Therefore, if the amount of steam generated by the steam boiler can be accurately predicted, the controllable parameters of the steam boiler can be adjusted so that the steam boiler generates the same amount of steam in the state of highest efficiency.
Most of domestic and foreign scholars have research on steam quantity prediction methods, and the existing methods mainly comprise a fuzzy neural network, a support vector machine, a neural network, a hybrid algorithm and the like. The prior prediction algorithm usually needs to select useful information in input characteristics according to the prior knowledge of professionals, and redundant information is removed. And the traditional model training mode has longer period and low accuracy.
Disclosure of Invention
The invention aims to provide a steam yield prediction method of a steam boiler based on data mining, and solves the problems that the existing prediction method is too dependent on experience, the training period is long and the accuracy is low.
In order to solve the technical problems, the invention adopts the following technical scheme: a method for predicting steam production of a steam boiler based on data mining is characterized by comprising the following steps:
s1, acquiring evaporation capacity production data of the gas-steam boiler and preprocessing the evaporation capacity production data, wherein the production data comprise ignition pressure, flameout pressure, steam temperature, boiler internal temperature, boiler water supply temperature, boiler water return pressure, natural gas flow, water supply pump rotating speed, water return pump rotating speed and steam flow; the preprocessing comprises data cleaning, data integration, data specification and data transformation;
s2, selecting parameters of the production data, and selecting parameters which have great influence on the prediction accuracy of the training data set; according to the characteristics of steam quantity prediction, a model method is adopted to screen characteristics;
s3, inputting the data after parameter screening into a training network, wherein the training network adopts a fusion model of self-adaptive weights, and the training network carries out data training on the screened data and constructs the fusion model of the self-adaptive weights to obtain a prediction model;
and S4, inputting the data of the test data set into the prediction model to obtain the prediction result of the steam amount.
A further technical scheme is that in the step S1, when the loss rate of the variables is greater than 80%, the coverage rate is low, the importance of the variables is predicted to be low through a steam generation principle, and the variables are directly deleted; fixed-value filling is replaced by-9999; and filling the missing data by using the mean value of the variable when the data accord with uniform distribution, and filling the missing data by using a median when the data have inclined distribution.
A further technical scheme is that in the step S2, a boosting-tree algorithm is adopted to calculate the importance index of each parameter, a parameter threshold is set, the parameter with the importance index larger than the parameter threshold is retained, and the parameter with the importance index smaller than the parameter threshold is deleted.
A further technical solution is that the specific training step in step S3 is:
s3-1, before model training, predicting by using ridge regression, and removing points of prediction residual errors in a training set, which are outside 3 sigma.
S3-2, in the model training stage, 85% of data of a training set are used for respectively carrying out primary training and training on a linear regression model, a support vector regression model and a random forest regression model, and training and testing of parameter tuning are needed for each model; in the random forest regression model, the following parameters are adjusted and optimized by using grid search: max _ features random forest allows a single decision tree to use the maximum number of features and the number of subtrees n _ estimators need to build; after the grid search, the parameters are respectively set as: sqrt and 200; using the rest 15% of the training set as a verification machine, and performing parameter adjustment processing on each model respectively;
and S3-3, in the self-adaptive model fusion stage, using the rest 15% of the training set as a verification set to verify various weighting methods, and determining the weights of the three models according to the minimum mean square error criterion.
The further technical scheme is that the production data obtained in the step S1 is selected from a Fuji oil/gas steam boiler, 4000 data are randomly selected to form a training data set, and the rest 1000 data form a testing data set.
Compared with the prior art, the invention has the beneficial effects that: the method applies a reasonable algorithm on the basis of the steam quantity production data, selects model parameters by using a boosting-tree algorithm, establishes a steam quantity prediction model by adopting a self-adaptive weight fusion model algorithm, selects corresponding process parameters, reduces input dimension and saves data training time. Compared with a prediction model of the steam volume constructed by a Neural Network (NN), a Support Vector Machine (SVM), a random forest tree and a k-nearest neighbor data mining algorithm, the prediction accuracy of the steam volume production prediction model constructed by the self-adaptive weight fusion model algorithm is higher.
Drawings
Fig. 1 is a structural diagram of a network model according to the present invention.
Fig. 2 is a graph of the effect of the fit of the steam generation.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Examples
A method for predicting the steam quantity of a gas steam boiler based on data mining comprises the following specific implementation steps:
s1, data acquisition
The steam production data is selected from a Foster fuel oil/gas steam boiler, and the production data comprises ignition pressure, flameout pressure, steam temperature, boiler internal temperature, boiler feed water temperature, boiler return water pressure, natural gas flow, feed water pump rotation speed, return water pump rotation speed and steam flow. All the data are collected by a Programmable Logic Controller (PLC) every 1 second, and 72-96 hours of operation data are collected.
The data set is randomly divided into two parts, wherein the first part of the data set 1 is provided with 4000 data points and used for training and developing a prediction model through a data mining algorithm, and the other part of the data set 2 is provided with 1000 data points and used for testing the prediction performance of the model derived from the data set 1.
And preprocessing the acquired data, including data cleaning, transformation and the like, and is used for deleting independent variables, normally converting the skewed data and standardizing the characteristic data.
The method for processing the missing value is mainly based on the distribution characteristics of the variables and the importance (information amount and prediction capability) of the variables, and specifically uses the following method:
deleting variables: if the deletion rate of the variable is high (more than 80%), the coverage rate is low, and the importance is low, the variable can be directly deleted.
Constant value filling: the substitution is carried out with-9999.
S2, parameter screening
The data set contains parameters of ignition pressure, flameout pressure, steam temperature, boiler internal temperature, boiler feed water temperature, boiler return water pressure, natural gas flow, feed water pump rotation speed, return water pump rotation speed and steam flow. Some of these parameters are important because they affect each other. Considering the steam quantity prediction problem, selecting less characteristic quantities can effectively reduce the model training period, and meanwhile, in order to improve the prediction precision, reducing the number of parameters by adopting a boosting-tree algorithm. And (3) calculating by using a boosting-tree algorithm to obtain the importance indexes of the parameters, wherein the importance indexes are shown in table 1.
Table 1 lists the importance indices for each input parameter. The threshold for the selection parameter is set to 0 and according to table 1 all parameters are more important than the threshold.
TABLE 1 importance index of input parameters
S3 model building of fusion model of self-adaptive weight
Linear regression, support vector regression and random forest regression are used as basic models, then a weighted fusion method is adopted for combination, namely, weighted average is carried out on prediction results of all models, and therefore large errors generated when a single model predicts a certain part of data can be avoided. The basic idea of weighted multi-model adaptive control is to adopt a 'divide-and-conquer' method, establish a plurality of local models and a plurality of corresponding local controllers offline, and fuse the control output of each local controller in an online weighted manner, thereby forming global control, and the method is an important method for realizing robust adaptive control. The optimal criterion employed by the adaptive algorithm is the least mean square error (LMS) criterion.
Before model training, a ridge regression is used for prediction, and abnormal samples (points with prediction residual errors outside 3 sigma) in a training set are removed.
The modeling process is divided into the following two phases: model training and adaptive model fusion.
In the model training stage, 85% of data of a training set are used for carrying out primary training on a linear regression model, a support vector regression model and a random forest regression model respectively. Meanwhile, it is noted that training and testing of parameter tuning are required for each model. In the random forest regression model, the following parameters are adjusted and optimized by using grid search: the max _ features random forest allows a single decision tree to use the maximum number of features and the number of subtrees that n _ estimators need to build. After the grid search, the parameters are respectively set as: sqrt and 200.
The remaining 15% of the training set was used as the verifier to separately parametrize each model.
In the self-adaptive model fusion stage, the rest 15% of the training set is used as a verification set to verify various weighting methods, and the weights of the three models are determined according to the least mean square error (LMS) criterion.
The toolbox functional linear regression, support vector regression, and random forest regression in Matlab 10.0(MathWorks, Inc.) were used to construct the prediction model. And inputting the data after parameter screening into a training network, and building the training network by adopting a self-adaptive weight fusion model. The method is used for constructing a basic model by using linear regression, support vector regression and random forest regression, and then combining the basic model by adopting a weighted fusion method, namely, weighted average is carried out on prediction results of each model, so that a large error generated when a certain part of data is predicted by a single model can be avoided.
S4 prediction model performance analysis
And obtaining observation and prediction data of the test data set based on the fusion model structure of the parameters and the self-adaptive weight. The result shows that the model established by the fusion model method of the self-adaptive weight can better predict the change of the steam quantity along with the time. In addition to the small variation between the observed and predicted values, the established model can clearly identify the majority of the steam production peaks.
The prediction accuracy of the data mining algorithm-derived model was evaluated using five indicators, Percent Error (PE), fractional deviation (FB), Root Mean Square Error (RMSE), normalized root mean square error (NMSE), and consistency Index (IA). The PE of the test data set was 0.10. The result shows that the predicted value is well matched with the measured value. FB is almost zero and NMSE is 0.015. These two measurements show that the difference between the predicted and observed values is small. The IA is very high, about 0.99. The predicted value and the measured value are well matched.
S5, verifying accuracy of the prediction model under the algorithm of the invention
A Neural Network (NN), a Support Vector Machine (SVM), a random forest tree and a k-nearest neighbor data mining algorithm are adopted to construct a prediction model of the steam amount. To obtain the best performing neural network, 200 networks were trained, with a maximum hidden unit of 30, and identity, logic, tanh, and exponent were chosen as the activation functions for hidden and output neurons. For support vector machine algorithms, the parameters of the kernel function: the capacity is between 10 and 20, the degree is between 1 and 5, and the gamma is between 0.2 and 1, so as to obtain the optimal parameter setting. The maximum number of iterations is set to 1000 to reduce test errors. For random forest trees, the number of predictors ranges from 2 to 6, and the number of trees ranges from 100 to 200. The seed of the random number generator is chosen between 1 and 5. For the k-nearest neighbor algorithm, the number of nearest neighbors is chosen as an odd number in the range of 1 to 15.
The result of fig. 2 shows that the fusion model of the adaptive weights has better prediction accuracy than the models established by other algorithms. In particular, the fractional deviation of the model derived from the fused model of adaptive weights is almost 0, which is significantly smaller than the values of other algorithms. The error percentage of the fusion model of the adaptive weights is minimal.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.
Claims (5)
1. A steam boiler steam yield prediction method based on data mining is characterized by comprising the following steps:
s1, acquiring evaporation capacity production data of the gas-steam boiler and preprocessing the evaporation capacity production data, wherein the production data comprise ignition pressure, flameout pressure, steam temperature, boiler internal temperature, boiler water supply temperature, boiler water return pressure, natural gas flow, water supply pump rotating speed, water return pump rotating speed and steam flow; the preprocessing comprises data cleaning, data integration, data specification and data transformation;
s2, selecting parameters of the production data, selecting important parameters which have great influence on the prediction accuracy of the training data set, and screening characteristics by adopting a model method according to the characteristics of steam quantity prediction;
s3, inputting the data after parameter screening into a training network, wherein the training network adopts a fusion model of self-adaptive weights, and the training network carries out data training on the screened data and constructs the fusion model of the self-adaptive weights to obtain a prediction model;
and S4, inputting the data of the test data set into the prediction model to obtain the prediction result of the steam amount.
2. A method for predicting steam production from a steam boiler based on data mining as claimed in claim 1, wherein: when the variable missing rate is greater than 80% in the data preprocessing in the step S1, the coverage rate is low, and the importance is judged to be low through the steam generation principle, and the variable is directly deleted; fixed-value filling is replaced by-9999; and filling the missing data by using the mean value of the variable when the data accord with uniform distribution, and filling the missing data by using a median when the data have inclined distribution.
3. A method for predicting steam production from a steam boiler based on data mining as claimed in claim 1, wherein: in step S2, a boosting-tree algorithm is used to calculate the importance index of each parameter, a parameter threshold is set, the parameter with the importance index greater than the parameter threshold is retained, and the parameter with the importance index less than the parameter threshold is deleted.
4. A method for predicting steam production from a steam boiler based on data mining as claimed in claim 1, wherein: the specific training step in step S3 is:
s3-1, before model training, predicting by using ridge regression, and removing points of prediction residual errors in a training set, which are outside 3 sigma; s3-2, in the model training stage, 85% of data of a training set are used for respectively carrying out primary training and training on a linear regression model, a support vector regression model and a random forest regression model, and training and testing of parameter tuning are needed for each model; in the random forest regression model, the following parameters are adjusted and optimized by using grid search: max _ features random forest allows a single decision tree to use the maximum number of features and the number of subtrees n _ estimators need to build; after the grid search, the parameters are respectively set as: sqrt and 200; using the rest 15% of the training set as a verification machine, and performing parameter adjustment processing on each model respectively; and S3-3, in the self-adaptive model fusion stage, using the rest 15% of the training set as a verification set to verify various weighting methods, and determining the weights of the three models according to the minimum mean square error criterion.
5. A method for predicting steam production from a steam boiler based on data mining as claimed in claim 1, wherein: the production data obtained in step S1 is selected from a fuji oil/gas steam boiler, and the obtained data is randomly obtained 4000 to form a training data set, and the remaining 1000 data are formed into a test data set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110317633.5A CN112906307A (en) | 2021-03-24 | 2021-03-24 | Steam yield prediction method of steam boiler based on data mining |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110317633.5A CN112906307A (en) | 2021-03-24 | 2021-03-24 | Steam yield prediction method of steam boiler based on data mining |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112906307A true CN112906307A (en) | 2021-06-04 |
Family
ID=76106277
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110317633.5A Pending CN112906307A (en) | 2021-03-24 | 2021-03-24 | Steam yield prediction method of steam boiler based on data mining |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112906307A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180087360A1 (en) * | 2016-09-26 | 2018-03-29 | International Business Machines Corporation | Controlling operation of a steam-assisted gravity drainage oil well system by adjusting controls based on forecast emulsion production |
| CN110222416A (en) * | 2019-06-05 | 2019-09-10 | 重庆邮电大学 | A kind of industrial steam amount prediction technique based on big data |
| CN110363289A (en) * | 2019-07-17 | 2019-10-22 | 山东浪潮人工智能研究院有限公司 | A kind of industrial steam amount prediction technique and device based on machine learning |
| CN110428053A (en) * | 2019-08-08 | 2019-11-08 | 东北大学 | A kind of steam produces the dynamic prediction method of consumption |
| US20200320237A1 (en) * | 2019-04-08 | 2020-10-08 | Doosan Heavy Industries & Construction Co., Ltd. | Apparatus and method for deriving boiler combustion model |
| CN112417764A (en) * | 2020-12-01 | 2021-02-26 | 江苏省特种设备安全监督检验研究院 | K nearest neighbor regression prediction method for boiler special equipment steam flow prediction |
-
2021
- 2021-03-24 CN CN202110317633.5A patent/CN112906307A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180087360A1 (en) * | 2016-09-26 | 2018-03-29 | International Business Machines Corporation | Controlling operation of a steam-assisted gravity drainage oil well system by adjusting controls based on forecast emulsion production |
| US20200320237A1 (en) * | 2019-04-08 | 2020-10-08 | Doosan Heavy Industries & Construction Co., Ltd. | Apparatus and method for deriving boiler combustion model |
| CN110222416A (en) * | 2019-06-05 | 2019-09-10 | 重庆邮电大学 | A kind of industrial steam amount prediction technique based on big data |
| CN110363289A (en) * | 2019-07-17 | 2019-10-22 | 山东浪潮人工智能研究院有限公司 | A kind of industrial steam amount prediction technique and device based on machine learning |
| CN110428053A (en) * | 2019-08-08 | 2019-11-08 | 东北大学 | A kind of steam produces the dynamic prediction method of consumption |
| CN112417764A (en) * | 2020-12-01 | 2021-02-26 | 江苏省特种设备安全监督检验研究院 | K nearest neighbor regression prediction method for boiler special equipment steam flow prediction |
Non-Patent Citations (3)
| Title |
|---|
| 李治等: "基于XGBoost的热电联产供热需求预测方法", 《电气自动化》 * |
| 李蔚等: "双重BP神经网络组合模型在实时数据预测中的应用", 《中国电机工程学报》 * |
| 青十五, 北京：机械工业出版社 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Shi et al. | Combustion optimization of ultra supercritical boiler based on artificial intelligence | |
| EP1921280B1 (en) | Systems and methods for multi-level optimizing control systems for boilers | |
| CN106094744A (en) | The determination method of thermoelectricity factory owner's operational factor desired value based on association rule mining | |
| CN107203687B (en) | Multi-target cooperative intelligent optimization control method for desulfurization process of absorption tower | |
| CN103235986A (en) | Operation and consumption optimization method based on boiler safety analysis | |
| CN108647808A (en) | A kind of manufacturing parameter Optimization Prediction method, apparatus, equipment and storage medium | |
| CN110458724A (en) | A method of based on neural network prediction fired power generating unit depth peak modulation capacity | |
| CN112377985B (en) | Heat supply unit peak regulation lower limit evaluation method and system based on feedback neural network | |
| CN115077640A (en) | Real-time prediction method for main steam flow in urban solid waste incineration process | |
| CN110207094A (en) | IQGA-SVR boiler heating surface fouling characteristics discrimination method based on principal component analysis | |
| CN111401652A (en) | Boiler optimization method and system based on CO online detection | |
| CN108647478B (en) | BP neural network-based SCR inlet smoke temperature online calculation method for cogeneration unit | |
| Chen et al. | Application of deep learning modelling of the optimal operation conditions of auxiliary equipment of combined cycle gas turbine power station | |
| US11475186B2 (en) | Apparatus and method for deriving boiler combustion model | |
| CN112906307A (en) | Steam yield prediction method of steam boiler based on data mining | |
| CN115437245A (en) | Boiler optimization control method based on combustion state prediction model | |
| CN107742000A (en) | Boiler combustion oxygen content modeling method | |
| CN115290218A (en) | Soft measurement method and system for wall temperature of boiler water wall of thermal generator set | |
| CN112348696B (en) | BP neural network-based heating unit peak regulation upper limit evaluation method and system | |
| Wang et al. | Data mining-based operation optimization of large coal-fired power plants | |
| CN115076716A (en) | Intelligent control method and system for porous medium burner | |
| CN114861126A (en) | Method and system for predicting stability in combustion adjustment process of gas turbine | |
| Yan et al. | Improved LS-SVM Boiler Combustion Model Based on Affinity Propagation | |
| CN112465198B (en) | Double-agent assisted search energy optimization method for small sample data of park | |
| CN111651938B (en) | Variable coal quality unit output prediction method based on thermodynamic calculation and big data |
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: 20210604 |
|
| RJ01 | Rejection of invention patent application after publication |