Method, device, storage medium and electronic equipment for predicting concentration of nitrogen oxides
By acquiring the boiler operating status time series and using probability and concentration prediction models to optimize and adjust the denitrification equipment, the problem of increased ammonia consumption caused by boiler load changes was solved, thus improving denitrification efficiency and economy.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHENHUA GUONENG ENERGY GRP
- Filing Date
- 2023-01-03
- Publication Date
- 2026-07-07
AI Technical Summary
In existing technologies, rapid changes in boiler load lead to increased ammonia consumption and severe ash accumulation in denitrification equipment, affecting denitrification efficiency and environmental and economic benefits.
By acquiring the operating time series of the target boiler, the change in nitrogen oxide concentration is predicted. Probabilistic prediction models and concentration prediction models are used to optimize and adjust the denitrification equipment to avoid increased ammonia consumption.
It improves the efficiency and environmental economy of denitrification equipment, reduces the amount of calculation, and ensures the accuracy of nitrogen oxide concentration prediction.
Smart Images
Figure CN116246727B_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of nitrogen oxide emission technology, and more specifically, to a method, apparatus, storage medium, and electronic device for predicting nitrogen oxide concentration. Background Technology
[0002] With economic development and social progress, the issue of sustainable environmental protection is receiving increasing attention. In industries such as coal chemical, steel, and power, most thermal power plants use ammonia injection into the boiler flue gas through denitrification equipment to ensure emissions meet standards. However, in actual boiler operation, rapid load changes and the passive follow-up control of denitrification lead to increased ammonia consumption and poor environmental economics. Furthermore, excessive ammonia leakage can react to form ammonium sulfate, causing severe ash accumulation on the denitrification catalyst and reducing denitrification efficiency; it can also cause frequent blockages in downstream equipment, jeopardizing the safe operation of the boiler. If the concentration level of nitrogen oxides in the boiler flue gas can be predicted, and considering the residence time of various parameters, an optimal adjustment method for each relevant parameter can be established to achieve economical and stable operation of environmental protection equipment. Summary of the Invention
[0003] The purpose of this disclosure is to provide a method, apparatus, storage medium, and electronic device for predicting nitrogen oxide concentrations, so as to optimize and adjust denitrification equipment based on the predicted nitrogen oxide concentrations.
[0004] To achieve the above objectives, in a first aspect, this disclosure provides a method for predicting nitrogen oxide concentrations, comprising:
[0005] Obtain the operating status of the target boiler within a first preset time period before the current moment to obtain the target time series;
[0006] Based on the target time series, determine whether the change in the nitrogen oxide concentration of the target boiler within a second preset time period after the current moment is greater than a preset change threshold.
[0007] If the change is greater than the preset change threshold, then the nitrogen oxide concentration of the target boiler after the second preset time period is predicted based on the target time series.
[0008] Optionally, determining whether the change in the nitrogen oxide concentration of the target boiler within a second preset time period after the current moment, based on the target time series, is greater than a preset change threshold includes:
[0009] The target time series is input into a pre-trained probability prediction model to obtain the probability that the change is greater than the preset change threshold.
[0010] Based on the probability, determine whether the amount of change is greater than the preset change threshold.
[0011] Optionally, predicting the nitrogen oxide concentration of the target boiler after the second preset time period based on the target time series includes:
[0012] The target time series is subjected to feature engineering processing;
[0013] Based on the target time series and the target time series obtained after feature engineering processing, the nitrogen oxide concentration of the target boiler is predicted after the second preset time period.
[0014] Optionally, the operating status includes excess air coefficient, furnace air volume, burnout air volume, coal volume, and furnace temperature;
[0015] The feature engineering process for the target time series includes:
[0016] Select target features from the operating state;
[0017] The target features are then subjected to feature processing.
[0018] Optionally, the feature processing includes at least one of differentiation, integration, normalization, and vectorization.
[0019] Optionally, predicting the nitrogen oxide concentration of the target boiler after the second preset time period based on the target time series and the target time series obtained after feature engineering processing includes:
[0020] The target time series is mapped to a first three-dimensional tensor, and the target time series obtained after feature engineering is mapped to a second three-dimensional tensor.
[0021] The first three-dimensional tensor and the second three-dimensional tensor are input into a pre-trained concentration prediction model to obtain the nitrogen oxide concentration of the target boiler after the second preset time.
[0022] Optionally, the concentration prediction model is trained in the following manner:
[0023] The operating status of the target boiler during a historical period is obtained to obtain a historical time series.
[0024] Obtain the actual nitrogen oxide concentration of the target boiler after the second preset time period at a historical moment, wherein the historical moment is the end time of the historical period;
[0025] The historical time series is subjected to feature engineering processing;
[0026] The historical time series is mapped to a third three-dimensional tensor, and the historical time series obtained after feature engineering is mapped to a fourth three-dimensional tensor.
[0027] The concentration prediction model is obtained by training the model using the third and fourth three-dimensional tensors as inputs and the actual nitrogen oxide concentration as the target output.
[0028] Secondly, this disclosure provides a nitrogen oxide concentration prediction device, comprising:
[0029] The first acquisition module is used to acquire the operating status of the target boiler within a first preset time period before the current moment, and obtain the target time series.
[0030] The first determining module is used to determine, based on the target time series, whether the change in the nitrogen oxide concentration of the target boiler within a second preset time period after the current moment is greater than a preset change threshold.
[0031] The second determining module is used to predict the nitrogen oxide concentration of the target boiler after the second preset time period based on the target time series if the change amount is greater than the preset change amount threshold.
[0032] Thirdly, this disclosure provides a computer-readable medium having a computer program stored thereon that, when executed by a processor, implements the steps of the method described in any of the first aspects.
[0033] Fourthly, this disclosure provides an electronic device, comprising: a memory having a computer program stored thereon; and a processor for executing the computer program in the memory to implement the steps of the method described in any of the first aspects.
[0034] In the above technical solution, the operating status of the target boiler within a first preset time period before the current moment is first obtained, resulting in a target time series. Then, based on the target time series, it is determined whether the change in the nitrogen oxide concentration of the target boiler within a second preset time period after the current moment exceeds a preset change threshold. If the change exceeds the preset change threshold, the nitrogen oxide concentration of the target boiler after the second preset time period is predicted based on the target time series. Thus, when the nitrogen oxide concentration changes drastically, it indicates a rapid change in the target boiler load, which may lead to increased ammonia consumption in the denitrification equipment. In this case, the nitrogen oxide concentration after the second preset time period after the current moment is predicted based on the target time series, allowing for optimization and adjustment of the denitrification equipment based on the predicted nitrogen oxide concentration. This avoids increased ammonia consumption in the denitrification equipment, improving denitrification efficiency and environmental economy. Furthermore, predicting the nitrogen oxide concentration based on the target time series takes into account the dynamic characteristics of the target boiler's operating status, ensuring the accuracy of the nitrogen oxide concentration prediction. Moreover, since the nitrogen oxide concentration prediction is only performed when drastic changes occur, rather than in real-time, the computational load is significantly reduced.
[0035] Other features and advantages of this disclosure will be described in detail in the following detailed description section. Attached Figure Description
[0036] The accompanying drawings are provided to further illustrate the present disclosure and form part of the specification. They are used together with the following detailed description to explain the present disclosure, but do not constitute a limitation thereof. In the drawings:
[0037] Figure 1 This is a flowchart of a nitrogen oxide concentration prediction method provided by an exemplary embodiment of this disclosure;
[0038] Figure 2 This is a block diagram of a nitrogen oxide concentration prediction device provided in an exemplary embodiment of this disclosure;
[0039] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an exemplary embodiment of the present disclosure;
[0040] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an exemplary embodiment of the present disclosure. Detailed Implementation
[0041] The specific embodiments of this disclosure will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit this disclosure.
[0042] This embodiment provides a method for predicting nitrogen oxide concentration. Figure 1 This is a flowchart of a nitrogen oxide concentration prediction method provided by an exemplary embodiment of this disclosure, such as... Figure 1 As shown, the method may include the following steps:
[0043] S101, obtain the operating status of the target boiler within the first preset time period before the current time to obtain the target time series.
[0044] As an example, operating status includes excess air coefficient, furnace air volume, burnout air volume, coal quantity, and furnace temperature.
[0045] The excess air coefficient refers to the ratio of the actual amount of air supplied for fuel combustion to the theoretical amount of air. The burnout air volume is the amount of hot air separately supplied to the upper part of the main burner in the furnace using a staged air supply method. The furnace air volume is the furnace purging air volume.
[0046] Among these factors, excess air coefficient, furnace air volume, burnout air volume, coal quantity, and furnace temperature are the factors affecting the change in nitrogen oxide concentration. Therefore, it is necessary to detect the excess air coefficient, furnace air volume, burnout air volume, coal quantity, and furnace temperature. For example, the oxygen concentration of the target boiler within a first preset time period before the current moment can be obtained through an oxygen sensor in the distributed control system. Based on these oxygen concentrations, the excess air coefficient of the target boiler within the first preset time period before the current moment can be calculated. The furnace air volume of the target boiler within the first preset time period before the current moment can be obtained through a first air volume sensor. The burnout air volume of the target boiler within the first preset time period before the current moment can be obtained through a second air volume sensor. The coal quantity of the target boiler within the first preset time period before the current moment can be obtained through a coal quantity sensor. The furnace temperature of the target boiler within the first preset time period before the current moment can be obtained through a temperature sensor. The target time series is a sequence of operating states arranged chronologically.
[0047] S102, Based on the target time series, determine whether the change in the nitrogen oxide concentration of the target boiler within a second preset time period after the current moment is greater than a preset change threshold.
[0048] The combustion of pulverized coal mainly produces the following nitrogen oxides: nitric oxide, nitrogen oxides, and nitrogen dioxide. Therefore, the nitrogen oxides in this disclosure can be: nitric oxide, nitrogen oxides, and nitrogen dioxide.
[0049] S103, if the change is greater than the preset change threshold, then predict the nitrogen oxide concentration of the target boiler after the second preset time period based on the target time series.
[0050] If the change exceeds a preset threshold, it indicates that the nitrogen oxide concentration of the target boiler changes drastically within a second preset time period after the current moment, meaning the target boiler load changes rapidly, potentially increasing ammonia consumption by the denitrification equipment. In this case, it is necessary to predict the nitrogen oxide concentration of the target boiler after the second preset time period based on the target time series, and optimize the denitrification equipment accordingly to avoid increasing ammonia consumption. If the change is not greater than the preset threshold, it indicates that the nitrogen oxide concentration of the target boiler does not change significantly within the second preset time period after the current moment, meaning the target boiler load changes slightly and will not increase ammonia consumption by the denitrification equipment. In this case, no optimization of the denitrification equipment is required, and therefore no prediction of the target boiler's nitrogen oxide concentration is needed. The process can then return to step S101 above.
[0051] In the above technical solution, the operating status of the target boiler within a first preset time period before the current moment is first obtained, resulting in a target time series. Then, based on the target time series, it is determined whether the change in the nitrogen oxide concentration of the target boiler within a second preset time period after the current moment exceeds a preset change threshold. If the change exceeds the preset change threshold, the nitrogen oxide concentration of the target boiler after the second preset time period is predicted based on the target time series. Thus, when the nitrogen oxide concentration changes drastically, it indicates a rapid change in the target boiler load, which may lead to increased ammonia consumption in the denitrification equipment. In this case, the nitrogen oxide concentration after the second preset time period after the current moment is predicted based on the target time series, allowing for optimization and adjustment of the denitrification equipment based on the predicted nitrogen oxide concentration. This avoids increased ammonia consumption in the denitrification equipment, improving denitrification efficiency and environmental economy. Furthermore, predicting the nitrogen oxide concentration based on the target time series takes into account the dynamic characteristics of the target boiler's operating status, ensuring the accuracy of the nitrogen oxide concentration prediction. Moreover, predicting the nitrogen oxide concentration only when drastic changes occur, rather than in real-time, significantly reduces the computational load.
[0052] In one possible embodiment, determining whether the change in the nitrogen oxide concentration of the target boiler within a second preset time period after the current moment is greater than a preset change threshold, based on the target time series, includes the following two steps:
[0053] Step 1: Input the target time series into the pre-trained probability prediction model to obtain the probability that the change is greater than the preset change threshold.
[0054] Among them, the probability prediction model can be a model based on a neural network used to predict the probability that the change in the nitrogen oxide concentration of the target boiler within a second preset time period after the current moment is greater than a preset change threshold. For example, the neural network can be a CNN-LSTM neural network composed of a convolutional neural network (CNN) and a long short-term memory artificial neural network (LSTM).
[0055] For example, a probabilistic prediction model can be trained in the following way:
[0056] The operating status of the target boiler within a historical period is obtained to obtain a historical time series. The probability of a sample where the change in nitrogen oxide concentration of the target boiler exceeds a preset threshold after a second preset time interval at a given historical moment is also obtained, where the historical moment is the end of the historical period. ； The above-mentioned probabilistic prediction model is obtained by training the model using historical time series as input and sample probabilities as the target output.
[0057] In one implementation, for each historical period under the same operating condition, the number of historical periods in which the change in nitrogen oxide concentration of the target boiler is greater than a preset change threshold after a second preset time period at the end of the corresponding period can be counted; then, the ratio of this number to the total number of each historical period under the same operating condition is determined as the sample probability that the change in nitrogen oxide concentration of the target boiler is greater than the preset change threshold after a second preset time period at the historical time.
[0058] Step 2: Based on the probability, determine whether the change is greater than the preset change threshold.
[0059] Specifically, if the predicted probability is greater than a preset probability threshold, then the change is determined to be greater than a preset change threshold. If the predicted probability is less than or equal to the preset probability threshold, then the change is determined to be less than or equal to a preset change threshold.
[0060] Based on the target time series, the nitrogen oxide concentration of the target boiler is predicted after a second preset time period, including the following two implementation methods:
[0061] In one possible embodiment, predicting the nitrogen oxide concentration of the target boiler after a second preset time period, based on the target time series, includes:
[0062] Based directly on the target time series, predict the nitrogen oxide concentration of the target boiler after a second preset time period.
[0063] Specifically, the target time series can be mapped to a first three-dimensional tensor, and then the first three-dimensional tensor can be input into the concentration prediction model to obtain the nitrogen oxide concentration of the target boiler after a second preset time period.
[0064] In another possible embodiment, predicting the nitrogen oxide concentration of the target boiler after a second preset time period based on the target time series includes the following two steps:
[0065] Step 1: Perform feature engineering on the target time series;
[0066] Feature engineering refers to using a series of engineered methods to select better data features from operational data. For example, target features can be formed by selecting certain types of data from the operational status.
[0067] Step 2: Based on the target time series and the target time series obtained after feature engineering processing, predict the nitrogen oxide concentration of the target boiler after the second preset time period.
[0068] As an example, based on the target time series and the target time series obtained after feature engineering processing, the nitrogen oxide concentration of the target boiler after a second preset time period is predicted, including:
[0069] The target time series is mapped to a first three-dimensional tensor, and the target time series obtained after feature engineering is mapped to a second three-dimensional tensor.
[0070] The first and second three-dimensional tensors are input into a pre-trained concentration prediction model to obtain the nitrogen oxide concentration of the target boiler after a second preset time period.
[0071] For example, the concentration prediction model can be a neural network-based model used to predict the nitrogen oxide concentration of a target boiler after a second preset time period. This neural network could be ConvLSTM, ConvGRU, C3D, etc. The input to the concentration prediction model is a three-dimensional tensor, therefore the target time series needs to be mapped to a three-dimensional tensor.
[0072] In the above technical solution, when predicting nitrogen oxide concentration, not only the target time series is referenced, but also the target time series obtained after feature engineering processing, thereby introducing richer feature information and improving the accuracy of nitrogen oxide concentration prediction.
[0073] As an example, feature engineering of a target time series may include the following steps:
[0074] Select target features from the running status;
[0075] Perform feature processing on the target features.
[0076] In one possible embodiment, the feature processing includes at least one of differentiation, integration, normalization, and vectorization.
[0077] For example, vectorization can be achieved through one-hot encoding.
[0078] In one possible embodiment, the concentration prediction model is trained in the following manner:
[0079] Obtain the operating status of the target boiler within a historical period to obtain the historical time series;
[0080] Obtain the actual nitrogen oxide concentration of the target boiler after a second preset time period at a historical moment, where the historical moment is the end time of the historical period;
[0081] Perform feature engineering on historical time series;
[0082] The historical time series is mapped to a third three-dimensional tensor, and the historical time series obtained after feature engineering is mapped to a fourth three-dimensional tensor.
[0083] The concentration prediction model is obtained by training the model using the third and fourth three-dimensional tensors as inputs and the actual nitrogen oxide concentration as the target output.
[0084] For example, the actual nitrogen oxide concentration can be measured by a nitrogen oxide concentration sensor. The process involves mapping the collected historical time series data to obtain a three-dimensional tensor, inputting this tensor into a concentration prediction model, and calculating the loss between the model's output and the actual nitrogen oxide concentration. Based on this loss, the model parameters are then adjusted using the gradient descent method.
[0085] It should be noted that the first preset duration, the second preset duration, and the preset change threshold can be set according to the actual application scenario, and this disclosure does not impose any restrictions here.
[0086] Figure 2 This is a block diagram of a nitrogen oxide concentration prediction device provided in an exemplary embodiment of the present disclosure. The device 10 includes:
[0087] The first acquisition module 500 is used to acquire the operating status of the target boiler within a first preset time period before the current moment, and obtain the target time series.
[0088] The first determining module 510 is used to determine, based on the target time series, whether the change in the nitrogen oxide concentration of the target boiler within a second preset time period after the current moment is greater than a preset change threshold.
[0089] The second determining module 520 is used to predict the nitrogen oxide concentration of the target boiler after the second preset time period based on the target time series if the change amount is greater than the preset change amount threshold.
[0090] In the above technical solution, the operating status of the target boiler within a first preset time period before the current moment is first obtained, resulting in a target time series. Then, based on the target time series, it is determined whether the change in the nitrogen oxide concentration of the target boiler within a second preset time period after the current moment exceeds a preset change threshold. If the change exceeds the preset change threshold, the nitrogen oxide concentration of the target boiler after the second preset time period is predicted based on the target time series. Thus, when the nitrogen oxide concentration changes drastically, it indicates a rapid change in the target boiler load, which may lead to increased ammonia consumption in the denitrification equipment. In this case, the nitrogen oxide concentration after the second preset time period after the current moment is predicted based on the target time series, allowing for optimization and adjustment of the denitrification equipment based on the predicted nitrogen oxide concentration. This avoids increased ammonia consumption in the denitrification equipment, improving denitrification efficiency and environmental economy. Furthermore, predicting the nitrogen oxide concentration based on the target time series takes into account the dynamic characteristics of the target boiler's operating status, ensuring the accuracy of the nitrogen oxide concentration prediction. Moreover, predicting the nitrogen oxide concentration only when drastic changes occur, rather than in real-time, significantly reduces the computational load.
[0091] Optionally, the first determining module 510 includes:
[0092] The first input submodule is used to input the target time series into a pre-trained probability prediction model to obtain the probability that the change is greater than the preset change threshold.
[0093] The determination submodule is used to determine whether the change amount is greater than the preset change amount threshold based on the probability.
[0094] Optionally, the second determining module 520 includes:
[0095] The first processing submodule is used to perform feature engineering processing on the target time series;
[0096] The prediction submodule is used to predict the nitrogen oxide concentration of the target boiler after the second preset time period based on the target time series and the target time series obtained after feature engineering processing.
[0097] Optionally, the operating status includes excess air coefficient, furnace air volume, burnout air volume, coal volume, and furnace temperature;
[0098] The first processing submodule includes:
[0099] A selection submodule is used to select target features from the running state;
[0100] The second processing submodule is used to perform feature processing on the target features.
[0101] Optionally, the feature processing includes at least one of differentiation, integration, normalization, and vectorization.
[0102] Optionally, the prediction submodule includes:
[0103] The mapping submodule is used to map the target time series into a first three-dimensional tensor and to map the target time series obtained after feature engineering into a second three-dimensional tensor.
[0104] The second input submodule is used to input the first three-dimensional tensor and the second three-dimensional tensor into a pre-trained concentration prediction model to obtain the nitrogen oxide concentration of the target boiler after the second preset time.
[0105] Optionally, the concentration prediction model is trained using a model training device, wherein the model training device includes:
[0106] The second acquisition module is used to acquire the operating status of the target boiler during a historical period to obtain a historical time series.
[0107] The third acquisition module is used to acquire the actual nitrogen oxide concentration of the target boiler after the second preset time period at a historical moment, wherein the historical moment is the end time of the historical period.
[0108] The processing module is used to perform feature engineering processing on the historical time series;
[0109] The mapping module is used to map the historical time series into a third three-dimensional tensor and to map the historical time series obtained after feature engineering into a fourth three-dimensional tensor.
[0110] The training module is used to train the concentration prediction model by using the third and fourth three-dimensional tensors as inputs to the concentration prediction model and the actual nitrogen oxide concentration as the target output of the concentration prediction model, so as to obtain the concentration prediction model.
[0111] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.
[0112] Figure 3 This is a block diagram illustrating an electronic device 700 according to an exemplary embodiment. For example... Figure 3As shown, the electronic device 700 may include: a first processor 701 and a first memory 702. The electronic device 700 may also include one or more of the following: a multimedia component 703, a first input / output (I / O) interface 704, and a communication component 705.
[0113] The first processor 701 controls the overall operation of the electronic device 700 to complete all or part of the steps in the nitrogen oxide concentration prediction method described above. The first memory 702 stores various types of data to support the operation of the electronic device 700. This data may include, for example, instructions for any application or method operating on the electronic device 700, and application-related data such as contact data, sent and received messages, pictures, audio, video, etc. The first memory 702 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. Multimedia component 703 may include a screen and an audio component. The screen may be, for example, a touchscreen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in the first memory 702 or transmitted via communication component 705. The audio component also includes at least one speaker for outputting audio signals. I / O interface 704 provides an interface between the first processor 701 and other interface modules, such as a keyboard, mouse, buttons, etc. These buttons may be virtual or physical buttons. Communication component 705 is used for wired or wireless communication between the electronic device 700 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, 4G, NB-IoT, eMTC, or other 5G technologies, or combinations thereof, is not limited here. Therefore, the corresponding communication component 705 may include: a Wi-Fi module, a Bluetooth module, an NFC module, etc.
[0114] In an exemplary embodiment, the electronic device 700 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the nitrogen oxide concentration prediction method described above.
[0115] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided, which, when executed by a processor, implement the steps of the nitrogen oxide concentration prediction method described above. For example, the computer-readable storage medium may be the first memory 702 including the program instructions described above, which may be executed by the first processor 701 of the electronic device 700 to complete the nitrogen oxide concentration prediction method described above.
[0116] Figure 4 This is a block diagram illustrating an electronic device 1900 according to an exemplary embodiment. For example, the electronic device 1900 may be provided as a server. (Refer to...) Figure 4 The electronic device 1900 includes a second processor 1922, which may be one or more, and a second memory 1932 for storing computer programs executable by the second processor 1922. The computer program stored in the second memory 1932 may include one or more modules, each corresponding to a set of instructions. Furthermore, the second processor 1922 may be configured to execute the computer program to perform the aforementioned nitrogen oxide concentration prediction method.
[0117] Additionally, the electronic device 1900 may also include a power supply component 1926 and a communication component 1950. The power supply component 1926 can be configured to perform power management of the electronic device 1900, and the communication component 1950 can be configured to enable communication of the electronic device 1900, such as wired or wireless communication. Furthermore, the electronic device 1900 may also include a second input / output (I / O) interface 1958. The electronic device 1900 can operate on an operating system, such as Windows Server, stored in a second memory 1932. TM Mac OS X TM UnixTM Linux TM etc.
[0118] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided, which, when executed by a processor, implement the steps of the nitrogen oxide concentration prediction method described above. For example, the non-transitory computer-readable storage medium may be the second memory 1932 including the program instructions, which may be executed by the second processor 1922 of the electronic device 1900 to complete the nitrogen oxide concentration prediction method described above.
[0119] In another exemplary embodiment, a computer program product is also provided, comprising a computer program executable by a programmable device, the computer program having a code portion for performing the above-described nitrogen oxide concentration prediction method when executed by the programmable device.
[0120] The preferred embodiments of this disclosure have been described in detail above with reference to the accompanying drawings. However, this disclosure is not limited to the specific details of the above embodiments. Within the scope of the technical concept of this disclosure, various simple modifications can be made to the technical solutions of this disclosure, and these simple modifications all fall within the protection scope of this disclosure.
[0121] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, this disclosure will not describe the various possible combinations separately.
[0122] Furthermore, various different embodiments of this disclosure can be combined in any way, as long as they do not violate the spirit of this disclosure, they should also be regarded as the content disclosed in this disclosure.
Claims
1. A method for predicting nitrogen oxide concentration, characterized in that, include: Obtain the operating status of the target boiler within a first preset time period before the current moment to obtain the target time series; Based on the target time series, determine whether the change in the nitrogen oxide concentration of the target boiler within a second preset time period after the current moment is greater than a preset change threshold. If the change is greater than the preset change threshold, then the nitrogen oxide concentration of the target boiler after the second preset time period is predicted based on the target time series. The step of determining whether the change in the nitrogen oxide concentration of the target boiler within a second preset time period after the current moment, based on the target time series, is greater than a preset change threshold includes: The target time series is input into a pre-trained probability prediction model to obtain the probability that the change is greater than the preset change threshold. Based on the probability, determine whether the amount of change is greater than the preset change threshold; The probability prediction model is trained in the following way: The operating status of the target boiler during a historical period is obtained to obtain a historical time series. Obtain the sample probability that the change in nitrogen oxide concentration of the target boiler is greater than a preset change threshold after a second preset time period at a historical moment, wherein the historical moment is the end time of the historical period. The probabilistic prediction model is obtained by training the model using the historical time series as input and the sample probability as the target output of the probabilistic prediction model. The step of predicting the nitrogen oxide concentration of the target boiler after the second preset time period based on the target time series includes: The target time series is subjected to feature engineering processing; Based on the target time series and the target time series obtained after feature engineering processing, predict the nitrogen oxide concentration of the target boiler after the second preset time period; The step of predicting the nitrogen oxide concentration of the target boiler after the second preset time period based on the target time series and the target time series obtained after feature engineering processing includes: The target time series is mapped to a first three-dimensional tensor, and the target time series obtained after feature engineering is mapped to a second three-dimensional tensor. The first three-dimensional tensor and the second three-dimensional tensor are input into a pre-trained concentration prediction model to obtain the nitrogen oxide concentration of the target boiler after the second preset time.
2. The method according to claim 1, characterized in that, The operating status includes excess air coefficient, furnace air volume, burnout air volume, coal volume, and furnace temperature; The feature engineering process for the target time series includes: Select target features from the operating states; The target features are then subjected to feature processing.
3. The method according to claim 2, characterized in that, The feature processing includes at least one of differentiation, integration, normalization, and vectorization.
4. The method according to claim 1, characterized in that, The concentration prediction model was trained in the following manner: The operating status of the target boiler during a historical period is obtained to obtain a historical time series. Obtain the actual nitrogen oxide concentration of the target boiler after the second preset time period at a historical moment, wherein the historical moment is the end time of the historical period; The historical time series is subjected to feature engineering processing; The historical time series is mapped to a third three-dimensional tensor, and the historical time series obtained after feature engineering is mapped to a fourth three-dimensional tensor. The concentration prediction model is obtained by training the model using the third and fourth three-dimensional tensors as inputs and the actual nitrogen oxide concentration as the target output.
5. A nitrogen oxide concentration prediction device, characterized in that, The device includes: The first acquisition module is used to acquire the operating status of the target boiler within a first preset time period before the current moment, and obtain the target time series. The first determining module is used to determine, based on the target time series, whether the change in the nitrogen oxide concentration of the target boiler within a second preset time period after the current moment is greater than a preset change threshold. The second determining module is used to predict the nitrogen oxide concentration of the target boiler after the second preset time period based on the target time series if the change amount is greater than the preset change amount threshold. The first determining module includes: The first input submodule is used to input the target time series into a pre-trained probability prediction model to obtain the probability that the change is greater than the preset change threshold. The probability prediction model is trained in the following way: The operating status of the target boiler during a historical period is obtained to obtain a historical time series. Obtain the sample probability that the change in nitrogen oxide concentration of the target boiler is greater than a preset change threshold after a second preset time period at a historical moment, wherein the historical moment is the end time of the historical period. The probabilistic prediction model is obtained by training the model using the historical time series as input and the sample probability as the target output of the probabilistic prediction model. A determination submodule is used to determine whether the amount of change is greater than the preset amount of change threshold based on the probability. The second determining module includes: The first processing submodule is used to perform feature engineering processing on the target time series; The prediction submodule is used to predict the nitrogen oxide concentration of the target boiler after the second preset time period based on the target time series and the target time series obtained after feature engineering processing. The mapping submodule is used to map the target time series into a first three-dimensional tensor and to map the target time series obtained after feature engineering into a second three-dimensional tensor. The second input submodule is used to input the first three-dimensional tensor and the second three-dimensional tensor into a pre-trained concentration prediction model to obtain the nitrogen oxide concentration of the target boiler after the second preset time.
6. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method described in any one of claims 1-4.
7. An electronic device, characterized in that, include: A memory on which computer programs are stored; A processor for executing the computer program in the memory to implement the steps of the method according to any one of claims 1-4.