CN113469449A - Optimizing control method and system for desulfurization system - Google Patents
Optimizing control method and system for desulfurization system Download PDFInfo
- Publication number
- CN113469449A CN113469449A CN202110791373.5A CN202110791373A CN113469449A CN 113469449 A CN113469449 A CN 113469449A CN 202110791373 A CN202110791373 A CN 202110791373A CN 113469449 A CN113469449 A CN 113469449A
- Authority
- CN
- China
- Prior art keywords
- desulfurization system
- parameter
- model
- desulfurization
- meta
- 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
- 238000006477 desulfuration reaction Methods 0.000 title claims abstract description 143
- 230000023556 desulfurization Effects 0.000 title claims abstract description 142
- 238000000034 method Methods 0.000 title claims abstract description 55
- 238000005457 optimization Methods 0.000 claims abstract description 54
- 238000012549 training Methods 0.000 claims abstract description 52
- 238000007781 pre-processing Methods 0.000 claims abstract description 12
- 238000004422 calculation algorithm Methods 0.000 claims description 29
- 238000012545 processing Methods 0.000 claims description 19
- 238000012795 verification Methods 0.000 claims description 16
- 238000012360 testing method Methods 0.000 claims description 12
- 238000012937 correction Methods 0.000 claims description 9
- 238000013528 artificial neural network Methods 0.000 claims description 7
- 238000003860 storage Methods 0.000 claims description 5
- 238000001914 filtration Methods 0.000 claims description 4
- 238000012216 screening Methods 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 3
- 238000010200 validation analysis Methods 0.000 claims 1
- 238000004088 simulation Methods 0.000 abstract description 11
- 239000007789 gas Substances 0.000 description 8
- 230000002068 genetic effect Effects 0.000 description 7
- 230000008569 process Effects 0.000 description 7
- 230000006870 function Effects 0.000 description 5
- 230000002829 reductive effect Effects 0.000 description 5
- 238000005096 rolling process Methods 0.000 description 4
- 238000005070 sampling Methods 0.000 description 4
- 210000002569 neuron Anatomy 0.000 description 3
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 description 2
- 235000019738 Limestone Nutrition 0.000 description 2
- NINIDFKCEFEMDL-UHFFFAOYSA-N Sulfur Chemical compound [S] NINIDFKCEFEMDL-UHFFFAOYSA-N 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000011049 filling Methods 0.000 description 2
- 239000003546 flue gas Substances 0.000 description 2
- 239000006028 limestone Substances 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 239000002002 slurry Substances 0.000 description 2
- 229910052717 sulfur Inorganic materials 0.000 description 2
- 239000011593 sulfur Substances 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000000739 chaotic effect Effects 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 239000003245 coal Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000010924 continuous production Methods 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 230000003009 desulfurizing effect Effects 0.000 description 1
- 238000003912 environmental pollution Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000000670 limiting effect Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000036961 partial effect Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000002922 simulated annealing Methods 0.000 description 1
- 238000012430 stability testing Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2474—Sequence data queries, e.g. querying versioned data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
Abstract
The embodiment of the invention provides an optimization control method of a desulfurization system and the optimization control system of the desulfurization system, and belongs to the field of thermal power generating units. The method comprises the following steps: collecting real-time operation parameters of the desulfurization system; preprocessing the real-time operation parameters to obtain training sample data; taking the training sample data as input data, and outputting a predicted time data sequence of the desulfurization system through a prediction meta-model; taking the predicted time data sequence as input data, and outputting an optimized adjustment scheme of the desulfurization system through an optimization meta-model; and adjusting the desulfurization system according to the optimized adjusting scheme. The scheme of the invention improves the accuracy of the simulation and prediction of the operation condition of the desulfurization system and ensures that the desulfurization system is continuously in the optimal operation state in the whole operation working interval.
Description
Technical Field
The invention relates to the technical field of thermal power generating units, in particular to an optimization control method of a desulfurization system and an optimization control system of the desulfurization system.
Background
The desulfurization system is an important environmental protection system in a thermal power system, and is mainly used for desulfurizing sulfur-containing substances in flue gas discharged from the tail end of a boiler system, so that the excessive sulfur in the discharged flue gas is avoided, and the environmental pollution is avoided. Therefore, the operation state of the desulfurization system directly affects the emission performance of the whole thermal power generating unit, and in order to avoid excessive emission, the desulfurization system is often required to continuously maintain the operation of redundancy performance, so that the desulfurization effect can be ensured, but the operation and maintenance cost of the system is also correspondingly improved. Nowadays, a plurality of optimizing control methods for a desulfurization system appear, namely, on the premise of ensuring desulfurization performance, energy waste is reduced as much as possible, so that the desulfurization system is always kept in an optimal state to operate. In order to realize the mode, the operation state of the desulfurization system has to be accurately acquired, the subsequent operation state prediction is carried out according to the current operation state, and then the advance adjustment is carried out according to the prediction result. The key point is how to accurately acquire the operating state of the desulfurization system and accurately predict the operating state of the system in the future.
The existing working condition simulation method directly uses real-time operation parameters of a desulfurization system to simulate the working conditions, the method can cause a large amount of invalid information integration, not only greatly increases the simulation time, but also causes larger deviation between the simulated working conditions and the actual conditions because the actual influence weight of each operation parameter cannot be obtained. In order to solve the problems of low accuracy of the simulation result of the operation condition of the desulfurization system, time delay and difficulty in prediction in the prior art, a new optimization control method of the desulfurization system is needed.
Disclosure of Invention
The embodiment of the invention aims to provide an optimization control method and system for a desulfurization system, so as to at least solve the problems of low accuracy, time delay and difficulty in prediction of the operation condition simulation result of the desulfurization system in the prior art.
In order to achieve the above object, a first aspect of the present invention provides an optimization control method for a desulfurization system, the method comprising: collecting real-time operation parameters of the desulfurization system; preprocessing the real-time operation parameters to obtain training sample data; taking the training sample data as input data, and outputting a predicted time data sequence of the desulfurization system through a prediction meta-model; taking the predicted time data sequence as input data, and outputting an optimized adjustment scheme of the desulfurization system through an optimization meta-model; and adjusting the desulfurization system according to the optimized adjusting scheme.
Optionally, the preprocessing the real-time operation parameters to obtain training sample data includes: filtering the real-time operation parameters, and reserving preset type information; processing the missing value of the preset type information; and performing unified standard processing on the preset type information after the missing value processing according to a preset standardization rule, and outputting the processed operation parameters as training sample data.
Optionally, the method further includes: obtaining the prediction meta-model, including: acquiring historical operating parameters of the desulfurization system, extracting features of the historical operating parameters, and taking the historical operating parameters after feature extraction as data sources; randomly segmenting the data source into a training data source, a verification data source and a test data source according to a preset proportion; performing prediction element model training in a preset neural network structure through the training data source; verifying and optimizing the prediction meta-model training through the verification data source; testing the prediction meta-model after verification and optimization adjustment through the test data source; and outputting the prediction meta-model after the test is finished.
Optionally, the segmentation ratio of the training data source, the verification data source and the test data source is 7: 1: 2.
Optionally, the method further includes: obtaining an optimizing meta-model, comprising: acquiring historical operating parameters of the desulfurization system; determining a correlation weight value between each characteristic attribute in the historical operating parameters according to a preset correlation rule algorithm; extracting a characteristic attribute combination with the maximum associated weight value, and taking a corresponding parameter type contained in the characteristic attribute combination as an influence parameter type influencing the desulfurization efficiency of the desulfurization system; and establishing an optimizing meta-model influencing the desulfurization efficiency of the desulfurization system according to the influencing parameter types.
Optionally, the outputting the optimized adjustment scheme of the desulfurization system through an optimization meta-model by using the predicted time data sequence as input data includes: screening out parameter types influencing the desulfurization efficiency of the desulfurization system from the predicted time data sequence according to the optimization meta-model to serve as target parameters; training to obtain an optimal parameter association combination by using the target parameters as input data and through an association model which is included by the optimizing meta model and influences the desulfurization efficiency of the desulfurization system; and determining the type of equipment needing to be adjusted and the corresponding adjustment amount as the optimized adjustment scheme according to the parameter type and the corresponding parameter value of the optimal parameter association combination.
Optionally, the types of parameters affecting the desulfurization efficiency of the desulfurization system include: SO at outlet end of desulfurization system2Concentration and liquid-gas ratio.
Optionally, the determining, according to the parameter type and the corresponding parameter value of the optimal parameter association combination, the type of the device to be adjusted and the corresponding adjustment amount includes: determining a basic operation parameter requirement value of the normal operation of the desulfurization system; comparing the parameter values of the parameters in the optimal parameter association combination with the basic operation parameter required values of the corresponding parameters; if the parameter value of the parameter in the optimal parameter association combination is smaller than the basic operation parameter required value of the corresponding parameter, taking the basic operation parameter required value as a correction target; and if the parameter value of the parameter in the optimal parameter association combination is larger than or equal to the basic operation parameter required value of the corresponding parameter, taking the parameter value of the parameter in the optimal parameter association combination as a correction target.
The second aspect of the present invention provides an optimization control method for a desulfurization system, the system comprising: the acquisition unit is used for acquiring real-time operation parameters of the desulfurization system; the processing unit is used for preprocessing the real-time operation parameters to obtain training sample data; a training unit to: taking the training sample data as input data, and outputting a predicted time data sequence of the desulfurization system through a prediction meta-model; taking the predicted time data sequence as input data, and outputting an optimized adjustment scheme of the desulfurization system through an optimization meta-model; and the execution unit is used for adjusting the desulfurization system according to the optimized adjustment scheme.
In another aspect, the present invention provides a computer-readable storage medium having instructions stored thereon, which, when executed on a computer, cause the computer to execute the above-mentioned optimization control method of a desulfurization system.
Through the technical scheme, the corresponding prediction model and the optimization model are determined according to the historical operating data of the corresponding desulfurization system, then the operating parameters of the desulfurization system are collected in real time, and the operating parameters are preprocessed. Invalid data are filtered, only valid data related to desulfurization performance are reserved, the data size is greatly reduced, and subsequent training time is shortened. Then, the operation singular number prediction is carried out according to the self development rule of the desulfurization system, whether the desulfurization system is in the best operation condition or not is judged according to the prediction result when a node is at a certain time in the future, and the system is adjusted in advance under the condition that the desulfurization system is not in the best operation condition. The system can be kept in the optimal operation state when the system is operated to the prediction time node.
The scheme of the invention improves the accuracy of the simulation and prediction of the operation condition of the desulfurization system and ensures that the desulfurization system is continuously in the optimal operation state in the whole operation working interval.
Additional features and advantages of embodiments of the invention will be set forth in the detailed description which follows.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the embodiments of the invention without limiting the embodiments of the invention. In the drawings:
FIG. 1 is a flow chart illustrating the steps of a method for optimizing a desulfurization system according to one embodiment of the present invention;
FIG. 2 is a flowchart illustrating steps for constructing a prediction meta-model according to one embodiment of the present invention;
FIG. 3 is a flowchart of the steps for constructing an optimizing meta-model according to an embodiment of the present invention;
fig. 4 is a system configuration diagram of an optimizing control system of a desulfurization system according to an embodiment of the present invention.
Description of the reference numerals
10-an acquisition unit; 20-a processing unit; 30-a training unit; 40-execution unit.
Detailed Description
The following detailed description of embodiments of the invention refers to the accompanying drawings. It should be understood that the detailed description and specific examples, while indicating the present invention, are given by way of illustration and explanation only, not limitation.
Fig. 4 is a system configuration diagram of an optimizing control system of a desulfurization system according to an embodiment of the present invention. As shown in fig. 4, an embodiment of the present invention provides an optimization control system of a desulfurization system, the system including: the acquisition unit 10 is used for acquiring real-time operation parameters of the desulfurization system; the processing unit 20 is configured to pre-process the real-time operation parameters to obtain training sample data; a training unit 30 for: taking the training sample data as input data, and outputting a predicted time data sequence of the desulfurization system through a prediction meta-model; taking the predicted time data sequence as input data, and outputting an optimized adjustment scheme of the desulfurization system through an optimization meta-model; and the execution unit 40 is used for adjusting the desulfurization system according to the optimized adjusting scheme.
FIG. 1 is a flow chart of a method for optimizing a desulfurization system according to an embodiment of the present invention. As shown in fig. 1, an embodiment of the present invention provides an optimization control method for a desulfurization system, the method including:
step S10: and collecting real-time operation parameters of the desulfurization system.
Specifically, the basic idea of the scheme of the invention is to predict the operation parameters of the system in a future period of time by monitoring the operation parameters of the desulfurization system in real time, and to adjust the system in advance when the desulfurization efficiency of the operation system is predicted to be reduced, so that on one hand, the desulfurization system can work at the highest desulfurization efficiency when the prediction time point is reached, and on the other hand, when partial equipment needs to be intervened when the system reaches the prediction time point, intervening equipment can be prepared in advance, and the timely response of the system is ensured. Therefore, during the operation of the desulfurization system, the operation parameters of the desulfurization system need to be collected in real time, and the operation parameters include the operation parameters of each device and the desulfurization effect embodied by the desulfurization system, such as the concentration of terminal SO2 and the feed-gas ratio. The current operation state of the desulfurization system can be accurately constructed through the parameters, the operation state of a period of time in the future can be deduced through the current operation state according to the operation rule of the desulfurization system, and the deduced operation state can be used as a reference basis for adjustment in advance. Preferably, the acquisition unit 10 is constructed in an original control unit of the desulfurization system, and the operation parameters are acquired through the control parameters of the desulfurization system and the sensor unit existing in the acquisition unit, so that the extension cost of the sensor unit is reduced, and the application range of the system is improved.
Step S20: and preprocessing the real-time operation parameters to obtain training sample data.
Specifically, the directly collected operation parameters include many invalid information, and the parameter formats of the devices are greatly different, such as SO2The concentration is a specific value, and the operation parameters of valves of all equipment of the desulfurization system only have an opening and closing state. In order to unify the data standards, the cluttered data needs to be processed into readable data with the unified standards, and then simulation training is performed. Preferably, a preset preprocessing rule is carried out for operating parameter processing, wherein the preprocessing rule comprises data filtering, data missing value processing and unified standard processing. Wherein, the data filtering is to filter the operation parameters according to the actual conditions affecting the desulfurization efficiency of the desulfurization system, remove the invalid information therein, and only reserve the information in advanceThe type of parameters which have an influence on the desulfurization efficiency is set. Part of parameters cannot be directly acquired through the acquisition module, unknown parameter guessing is needed to be carried out according to the existing parameters, and missing value filling is needed to be carried out through missing value processing. On one hand, filling the operation parameters needing to be obtained by inference, and on the other hand, supplementing missing data in the operation parameters according to the parameter rule. Finally, the complete supplementary operating parameters are subjected to unified standard processing, although the retained data can affect the desulfurization efficiency, the influence degrees of the operating parameters are different, in order to avoid amplifying the influence weight of the operating parameters with small influence degrees, weight optimization needs to be carried out according to the actual conditions of the operating parameters, and the final operating parameters are ensured to be closer to the actual conditions.
Step S30: and taking the training sample data as input data, and outputting a predicted time data sequence of the desulfurization system through a prediction meta-model.
Specifically, the operation parameters after the preprocessing are used as training samples and input into a preset prediction meta-model, and the prediction meta-model can output a prediction time sequence of a period of time in the future according to a preset model rule, wherein the prediction time sequence comprises a time point of the period of time in the future and a specific numerical value of the operation parameters corresponding to the time point. The prediction time data sequence is presumed according to the previous operation rule of the corresponding desulfurization system, the existing rule can be directly used for prediction in the subsequent real-time monitoring process, but a prediction meta-model needs to be constructed when the system is constructed.
The prediction meta-model of the desulfurization system needs to be constructed by depending on the operation parameters influencing the desulfurization efficiency, and the operation state prediction in a future period of time is carried out by the existing operation parameters and prediction method. The conventional time sequence prediction method comprises a prediction model based on a statistical learning method, a prediction model based on machine learning and a deep learning algorithm model, wherein the statistical learning method is developed at the earliest, but the method is usually good at processing linear or weak nonlinear objects, and an accurate mathematical model is usually required to be established for describing the running state of the objects, which is not suitable for the production process of the environment-friendly island with strong coupling and nonlinearity. Compared with a statistical learning method, the machine learning prediction model and the deep learning prediction model directly take operation data as objects, and have stronger processing capability on complex nonlinear problems, and the prediction models comprise: the method comprises a moving average difference autoregressive algorithm, an online support vector regression recurrent neural network algorithm, a gate cycle unit algorithm improved on the basis of RNN, and a long-time and short-time memory network. The corresponding applicable ranges of the methods are different, because the operation load and the coal burning efficiency of the thermal power generating unit are dynamically changed, the operation parameters of the whole desulfurization system are changed randomly, and the method has strong nonlinearity. Therefore, preferably, the prediction is performed by using an LSTM (long-short time memory network) algorithm, and the LSTM algorithm is suitable for application environments with high coupling, nonlinearity, high dimensionality and long time lag, which is most appropriate to the actual operation state of the desulfurization system. Specifically, as shown in fig. 2, the construction of the prediction meta-model includes:
step S301: and performing prediction meta-model training data preparation.
Specifically, according to target variables of a prediction model, feature extraction is carried out on dimension variables by using a random forest algorithm, preprocessed historical data are selected as a data source of the model, the data source is randomly segmented, 70% of the segmented data are used as training data sources, and the segmented data are used as regular screening data and are used as data samples of the training model. And then 10% of data is cut out to be used as a verification data source, if the same data is directly used for training model verification, the problem of the model cannot be found, and through random cutting, part of historical data is used as known data for reverse verification of the model, so that whether the model obtained through training meets the actual condition can be verified. The method comprises the steps of taking the front-end time data of a verification data source as input data, obtaining time sequence data through training, comparing the time sequence data obtained through training with corresponding rear-end time data in the verification data source, if the difference value of the time sequence data and the corresponding rear-end time data is within a preset range, indicating that a trained prediction meta-model accords with the actual condition, and if the difference value is within the preset range, indicating that the model has a problem and needing to be trained again. The training model after passing the verification can ensure that the model meets the actual condition, but the stability of the model can not be verified only by a small part of data, so that 30% of data are preferably cut out as test data. For subsequent stability testing of the model.
Step S302: and constructing a prediction meta-model.
Specifically, before training of the model, firstly a neural network structure is required to be built, the neural network structure comprises an input layer, an LSTM layer and a full connection layer, training data are input into the network, each neuron in the network calculates a result and outputs the result to a next layer of neurons until the last layer of output outputs the result, in addition, a BP algorithm is used for feeding back the calculation error of each layer to the previous layer, the model corrects the participation weight of each neuron according to the calculation error, and the iteration is repeated until the output error of the model is converged, so that the training of the model is completed. And importing the screened training sample data into the constructed neural network structure for training until convergence is completed, and outputting a prediction meta-model. And then sequentially using a verification data source and a test data source to perform model verification and optimization until a stable and accurate prediction meta-model is output.
In a possible implementation mode, when the prediction meta-model is used for carrying out the simulation of the predicted time data sequence of the desulfurization system, the data in the latest time window is obtained in the form of a sliding window, data preprocessing operations such as cleaning, null value processing and data standardization are carried out, the processed data is input into the prediction model, the model outputs the prediction result in the time window, and the model sequentially outputs the predicted values of the target variable along with the sliding of the time window to finally form the time sequence value of the target variable.
Step S40: and taking the predicted time data sequence as input data, and outputting an optimized adjustment scheme of the desulfurization system through an optimization meta-model.
Specifically, the predicted time data sequence of the thermal power generating unit is obtained according to the steps, namely, the future development trend influencing the desulfurization efficiency parameters is found, however, the influence degrees of the efficiency on the desulfurization system are different, and if all the predicted parameters are used for performing the efficiency simulation of the desulfurization system, the data volume is too large, the real simulation of the efficiency of the desulfurization system cannot be guaranteed, the efficiency simulation time delay of the desulfurization system is too high, and too much time is wasted for mastering the future desulfurization efficiency in advance. In order to solve the problems, data types in the operation parameters are randomly combined, and on one hand, the influence weight of the combination of the parameter types is judged, and on the other hand, the influence weight of the parameter types on the desulfurization efficiency is judged. And screening out the parameter combination with the largest influence weight, and then accurately and quickly calculating the actual condition of the desulfurization efficiency only through the parameter combination. Therefore, as shown in fig. 3, the method specifically comprises the following steps:
step S401: and mining the association interval relation of the optimal parameter combination influencing the desulfurization efficiency.
Specifically, association rule analysis refers to finding an association relationship between each feature attribute in a data set by using an association rule algorithm based on a statistical rule presented by sample data, so as to obtain an association rule among features. From the rules derived by the algorithm, information of one attribute can be inferred from information of another attribute. The confidence of the rule has different values, and when the confidence reaches a certain threshold, the rule can be considered to be established. Common association rule algorithms (such as Apriori algorithm) include two concepts of support and confidence. The general association rule algorithm mainly takes the minimum support degree and the minimum confidence degree as constraint conditions and outputs rules among the features meeting the minimum support degree and the minimum confidence degree. The support and confidence of a rule are shown in the following formula:
in the real-time data acquisition process of the desulfurization system, data is continuous data formed at certain sampling intervals, parameters of each measuring point have certain data fluctuation instead of being stabilized at a fixed constant, so that discretization processing needs to be carried out on the data, continuous production data are converted into parameter intervals, and in the actual application process, the interval where the parameters are located is taken as a research object, so that the association rule which is more in line with actual application can be obtained.
Preferably, the rule mining algorithm uses Apriori principle to search for frequent item sets. Apriori states that if an item set is an infrequent item set, then all of its supersets are also infrequent. The time spent on searching for a frequent item set can be reduced by utilizing the principle. Establishing liquid-gas ratio and outlet SO based on the above process2The correlation model of the concentration is that after the data of two dimensional variables are discretized, the liquid-gas ratio and the outlet SO are set2And searching for a rule in the sample data based on Ariori, evaluating the output result of the model, and readjusting the parameters if the obtained rule is less or no, SO as to obtain a liquid-gas ratio and an outlet SO2And (4) final correlation of concentration.
Step S402: and establishing an optimizing meta-model.
Specifically, the intelligent optimization meta-model aims to solve the optimization problem in the complex problem, that is, for any solution space, the combination optimization of finding the optimal solution to make the objective function value optimal is significant in the optimization control, for example, in the desulfurization efficiency optimization control, the desulfurization efficiency and the outlet SO are found2The optimum combination of concentrations to achieve optimum control of desulfurization efficiency. There are many common intelligent optimization algorithms, and the classical algorithm includes: linear programming and dynamic programming; the guiding search method comprises the following steps: simulated annealing, genetic algorithms, and the like; the system dynamic evolution algorithm comprises the following steps: neural networks, chaotic searches, and the like. In the environmental island optimization control process, a genetic algorithm is taken as an example to illustrate the specific algorithm process of the genetic algorithm and how to realize intelligent optimization. Genetic Algorithms (GA) target all individuals in a population and use randomization techniques to guide an efficient search of an encoded parameter space. Wherein the selection, crossover and mutation constitute genetic operations of the genetic algorithm; parameter coding, initial population setting, fitness function design and heredityFive elements of operation design and control parameter setting form the core content of the genetic algorithm. And establishing an optimizing meta-model of the desulfurization system according to the rules.
Preferably, when the optimization meta-model is executed, the optimization step is performed in a rolling optimization manner, and model predictive control is an optimization control algorithm which determines a future control action through optimization of a certain performance index, namely, optimal value search is performed on the premise of ensuring the basic operation parameter requirement of normal operation of the desulfurization system. Determining the basic operation parameter requirement of the normal operation of the desulfurization system; comparing the parameters in the optimal parameter association combination with the basic operation parameter requirements of the corresponding parameters; if the comparison parameter is smaller than the basic operation parameter requirement of the corresponding parameter, taking the basic operation parameter as a correction target; and if the comparison parameter is larger than the basic operation parameter requirement of the corresponding parameter, taking the corresponding parameter in the optimal parameter association combination as a correction target. In addition, the optimization in MPC is a rolling optimization in a limited period, that is, at each sampling moment, the optimization index is only designed to be optimized from the moment to the future limited time, and as the time goes by the next sampling moment, the optimization period also goes forward, so that the rolling optimization in MPC is repeatedly operated on line, and is not completed off line once.
Step S50: and carrying out advanced adjustment on the desulfurization system according to the optimized adjustment scheme.
Specifically, after an optimal parameter combination is found according to the optimizing meta-model, according to a prediction parameter combination sequence, a prediction parameter with a numerical value larger than a preset threshold value in a corresponding predicted parameter item and the optimal parameter combination is corrected to the optimal parameter. And adjusting corresponding system equipment, and adjusting the optimal value of the operation parameter of the system when the predicted time point is ensured, so as to realize the advanced control of the optimization of the system.
Preferably, the model is corrected in real time during the subsequent use of the model, because as the system ages and adjusts, the training model applicable to the current system may not be able to complete the fitting of the subsequent system. To avoid the further the deviation of the model training results, it is preferable to operate on the modelIn the use process, model verification and optimization are carried out according to real-time operation parameters, and the model is guaranteed to be always suitable for the current system condition. In the actual prediction control, after a future control strategy is determined through optimization, the control at the current moment is not implemented immediately, but the actual output of the last control is detected firstly, and the MPC prediction is corrected by using the real-time information, on one hand, the correction can predict the future error and compensate the error on the basis of keeping the prediction model continuously, and on the other hand, the model can be directly corrected according to the actual error. Thus, during optimal control of the desulfurization system, the model is based on the last SO2Prediction result and actual SO of concentration prediction meta-model2And (3) concentration error data, establishing an error correction model (for example, predicting the error based on an XgBoost algorithm), and compensating the output result of the next meta-model by using the error correction model. Based on this, the control optimization in MPC is not only based on models, but also uses the actual information fed back, thus forming a closed loop control.
In a possible embodiment, the predicted time series value in a certain time step in the future is as follows through the methodLiquid-gas ratio x at present time1Desulfurization efficiency x2The corresponding obtained optimizing meta-model is as follows:
x2=g(x1,Rx),x1∈(0,1)
Qx,Rx∈R
wherein f (x) represents a mapping function of the prediction meta-model, and g (x) represents a liquid-gas ratioDesulfurization efficiency, gamma represents the outlet SO2Concentration red line value; qx,RxRespectively representing the dimensional variable matrixes of the implicit function and the relational function except for the liquid-gas ratio and the desulfurization efficiency. The optimal solution of the model is calculated by means of an intelligent optimization meta-model, such as a GA algorithm. Then, in the control process, the density of the limestone slurry needs to be maintained at 1250mg/Nm3Therefore, the minimum sum of the square of the residual error between the object output and the numerical value 1250 on future sampling points can be selected as an optimal control target, and the control model is continuously subjected to rolling optimization along with the continuous arrival of new data to form a control closed loop, so that the density of limestone slurry is maintained at 1250 +/-10 mg/Nm3。
Embodiments of the present invention also provide a computer-readable storage medium, which stores instructions that, when executed on a computer, cause the computer to execute the above-mentioned optimization control method for a desulfurization system.
Those skilled in the art will appreciate that all or part of the steps in the method for implementing the above embodiments may be implemented by a program, which is stored in a storage medium and includes several instructions to enable a single chip, a chip, or a processor (processor) to execute all or part of the steps in the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
While the embodiments of the present invention have been described in detail with reference to the accompanying drawings, the embodiments of the present invention are not limited to the details of the above embodiments, and various simple modifications can be made to the technical solution of the embodiments of the present invention within the technical idea of the embodiments of the present invention, and the simple modifications are within the scope of the embodiments of the present invention. It should be noted that the various features described in the above embodiments may be combined in any suitable manner without departing from the scope of the invention. In order to avoid unnecessary repetition, the embodiments of the present invention will not be described separately for the various possible combinations.
In addition, any combination of the various embodiments of the present invention is also possible, and the same should be considered as disclosed in the embodiments of the present invention as long as it does not depart from the spirit of the embodiments of the present invention.
Claims (10)
1. A method for optimizing control of a desulfurization system, the method comprising:
collecting real-time operation parameters of the desulfurization system;
preprocessing the real-time operation parameters to obtain training sample data;
taking the training sample data as input data, and outputting a predicted time data sequence of the desulfurization system through a prediction meta-model;
taking the predicted time data sequence as input data, and outputting an optimized adjustment scheme of the desulfurization system through an optimization meta-model;
and adjusting the desulfurization system according to the optimized adjusting scheme.
2. The method of claim 1, wherein preprocessing the real-time operating parameters to obtain training sample data comprises:
filtering the real-time operation parameters, and reserving preset type information;
processing the missing value of the preset type information;
and performing unified standard processing on the preset type information after the missing value processing according to a preset standardization rule, and outputting the processed operation parameters as training sample data.
3. The method of claim 1, further comprising: obtaining the prediction meta-model, including:
acquiring historical operating parameters of the desulfurization system, extracting features of the historical operating parameters, and taking the historical operating parameters after feature extraction as data sources;
randomly segmenting the data source into a training data source, a verification data source and a test data source according to a preset proportion;
performing prediction element model training in a preset neural network structure through the training data source;
verifying and optimizing the prediction meta-model training through the verification data source;
testing the prediction meta-model after verification and optimization adjustment through the test data source;
and outputting the prediction meta-model after the test is finished.
4. The method of claim 3, wherein the training data source, the validation data source, and the test data source are sliced at a 7: 1: 2 ratio.
5. The method of claim 1, further comprising: obtaining an optimizing meta-model, comprising:
acquiring historical operating parameters of the desulfurization system;
determining a correlation weight value between each characteristic attribute in the historical operating parameters according to a preset correlation rule algorithm;
extracting a characteristic attribute combination with the maximum associated weight value, and taking a corresponding parameter type contained in the characteristic attribute combination as an influence parameter type influencing the desulfurization efficiency of the desulfurization system;
and establishing an optimizing meta-model influencing the desulfurization efficiency of the desulfurization system according to the influencing parameter types.
6. The method of claim 5, wherein outputting the optimized tuning of the desulfurization system via an optimization meta-model using the predicted time data sequence as input data comprises:
screening out parameter types influencing the desulfurization efficiency of the desulfurization system from the predicted time data sequence according to the optimization meta-model to serve as target parameters;
training to obtain an optimal parameter association combination by using the target parameters as input data and through an association model which is included by the optimizing meta model and influences the desulfurization efficiency of the desulfurization system;
and determining the type of equipment needing to be adjusted and the corresponding adjustment amount as the optimized adjustment scheme according to the parameter type and the corresponding parameter value of the optimal parameter association combination.
7. The method of claim 6, wherein the types of parameters that affect the desulfurization efficiency of the desulfurization system include:
SO at outlet end of desulfurization system2Concentration and liquid-gas ratio.
8. The method according to claim 6, wherein the determining the type of the device to be adjusted and the corresponding adjustment amount according to the parameter type and the corresponding parameter value of the optimal parameter association combination comprises:
determining a basic operation parameter requirement value of the normal operation of the desulfurization system;
comparing the parameter values of the parameters in the optimal parameter association combination with the basic operation parameter required values of the corresponding parameters;
if the parameter value of the parameter in the optimal parameter association combination is smaller than the basic operation parameter required value of the corresponding parameter, taking the basic operation parameter required value as a correction target;
and if the parameter value of the parameter in the optimal parameter association combination is larger than or equal to the basic operation parameter required value of the corresponding parameter, taking the parameter value of the parameter in the optimal parameter association combination as a correction target.
9. An optimization control system for a desulfurization system, the system comprising:
the acquisition unit is used for acquiring real-time operation parameters of the desulfurization system;
the processing unit is used for preprocessing the real-time operation parameters to obtain training sample data;
a training unit to:
taking the training sample data as input data, and outputting a predicted time data sequence of the desulfurization system through a prediction meta-model;
taking the predicted time data sequence as input data, and outputting an optimized adjustment scheme of the desulfurization system through an optimization meta-model;
and the execution unit is used for adjusting the desulfurization system according to the optimized adjustment scheme.
10. A computer-readable storage medium having instructions stored thereon, which when executed on a computer, cause the computer to execute the optimization control method of the desulfurization system according to any one of claims 1 to 8.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110791373.5A CN113469449A (en) | 2021-07-13 | 2021-07-13 | Optimizing control method and system for desulfurization system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110791373.5A CN113469449A (en) | 2021-07-13 | 2021-07-13 | Optimizing control method and system for desulfurization system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113469449A true CN113469449A (en) | 2021-10-01 |
Family
ID=77879982
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110791373.5A Pending CN113469449A (en) | 2021-07-13 | 2021-07-13 | Optimizing control method and system for desulfurization system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113469449A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114177747A (en) * | 2021-12-02 | 2022-03-15 | 昆岳互联环境技术(江苏)有限公司 | Flue gas desulfurization sulfur dioxide concentration prediction method based on machine learning algorithm |
CN114815628A (en) * | 2022-06-30 | 2022-07-29 | 微山县大卫新型建材有限公司 | Desulfurization system control model and desulfurization system control method |
CN115094193A (en) * | 2022-06-27 | 2022-09-23 | 中冶华天南京工程技术有限公司 | Intelligent molten iron pretreatment desulfurization system based on data mining |
CN115400571A (en) * | 2022-08-31 | 2022-11-29 | 华能湖南岳阳发电有限责任公司 | Automatic slurry supply method and device for desulfurization absorption tower |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102693451A (en) * | 2012-06-14 | 2012-09-26 | 东北电力大学 | Method for predicting ammonia process flue gas desulphurization efficiency based on multiple parameters |
CN109034260A (en) * | 2018-08-07 | 2018-12-18 | 东南大学 | Desulfurizing tower oxidation fan fault diagnosis system and method based on Statistics and intelligent optimizing |
CN110033141A (en) * | 2019-04-22 | 2019-07-19 | 大唐环境产业集团股份有限公司 | A kind of method for building up of desulphurization system operating condition database |
CN110263988A (en) * | 2019-06-06 | 2019-09-20 | 东南大学 | A kind of data run optimization method based on power plant desulphurization system |
-
2021
- 2021-07-13 CN CN202110791373.5A patent/CN113469449A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102693451A (en) * | 2012-06-14 | 2012-09-26 | 东北电力大学 | Method for predicting ammonia process flue gas desulphurization efficiency based on multiple parameters |
CN109034260A (en) * | 2018-08-07 | 2018-12-18 | 东南大学 | Desulfurizing tower oxidation fan fault diagnosis system and method based on Statistics and intelligent optimizing |
CN110033141A (en) * | 2019-04-22 | 2019-07-19 | 大唐环境产业集团股份有限公司 | A kind of method for building up of desulphurization system operating condition database |
CN110263988A (en) * | 2019-06-06 | 2019-09-20 | 东南大学 | A kind of data run optimization method based on power plant desulphurization system |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114177747A (en) * | 2021-12-02 | 2022-03-15 | 昆岳互联环境技术(江苏)有限公司 | Flue gas desulfurization sulfur dioxide concentration prediction method based on machine learning algorithm |
CN115094193A (en) * | 2022-06-27 | 2022-09-23 | 中冶华天南京工程技术有限公司 | Intelligent molten iron pretreatment desulfurization system based on data mining |
CN114815628A (en) * | 2022-06-30 | 2022-07-29 | 微山县大卫新型建材有限公司 | Desulfurization system control model and desulfurization system control method |
CN114815628B (en) * | 2022-06-30 | 2022-09-02 | 微山县大卫新型建材有限公司 | Desulfurization system control model and desulfurization system control method |
CN115400571A (en) * | 2022-08-31 | 2022-11-29 | 华能湖南岳阳发电有限责任公司 | Automatic slurry supply method and device for desulfurization absorption tower |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113469449A (en) | Optimizing control method and system for desulfurization system | |
CN109816221B (en) | Project risk decision method, apparatus, computer device and storage medium | |
CN110245802B (en) | Cigarette empty-head rate prediction method and system based on improved gradient lifting decision tree | |
CN111148118A (en) | Flow prediction and carrier turn-off method and system based on time sequence | |
CN109508818B (en) | Online NOx prediction method based on LSSVM | |
CN111126658A (en) | Coal mine gas prediction method based on deep learning | |
CN111738477B (en) | Power grid new energy consumption capability prediction method based on deep feature combination | |
CN110782658A (en) | Traffic prediction method based on LightGBM algorithm | |
CN111144552B (en) | Multi-index grain quality prediction method and device | |
Ning et al. | GA-BP air quality evaluation method based on fuzzy theory. | |
CN111461463A (en) | Short-term load prediction method, system and equipment based on TCN-BP | |
CN112232373A (en) | Method and system for predicting carbon content of boiler fly ash | |
CN111476274B (en) | Big data predictive analysis method, system, device and storage medium | |
CN114266384A (en) | Method and device for rapidly predicting typhoon path based on fusion model | |
CN112184412A (en) | Modeling method, device, medium and electronic equipment of credit rating card model | |
CN113449919B (en) | Power consumption prediction method and system based on feature and trend perception | |
CN114358116A (en) | Oil-immersed transformer fault diagnosis method and system and readable storage medium | |
CN113111588B (en) | NO of gas turbine X Emission concentration prediction method and device | |
CN114298411A (en) | GDP prediction method based on N-BEATS | |
CN113962158A (en) | Road roller construction process knowledge base construction method and system based on improved decision tree | |
CN113642784A (en) | Wind power ultra-short term prediction method considering fan state | |
CN113759709A (en) | Method and device for training strategy model, electronic equipment and storage medium | |
CN112348275A (en) | Regional ecological environment change prediction method based on online incremental learning | |
CN117744704B (en) | Flue gas pollution source acquisition monitoring system, method and readable storage medium | |
CN116957166B (en) | Tunnel traffic condition prediction method and system based on Hongmon system |
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 |