Method and device for monitoring blending combustion condition of coal and ammonia
By acquiring flame images in real time and performing image segmentation and feature extraction, combined with model training and loss function evaluation, and dynamically adjusting strategies to optimize the model, intelligent and automated monitoring of the coal and ammonia co-firing process is achieved. This solves the problem of difficult monitoring of combustion status in traditional methods and improves the accuracy and safety of combustion status.
Patent Information
- Application Number
- CN202511076862.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-01
- Publication Date
- 2025-11-14
AI Technical Summary
Traditional methods are insufficient for real-time monitoring and accurate assessment of the combustion status during the co-firing of coal and ammonia, leading to pollutants and harmful gases posing a threat to the environment and human health.
A method for monitoring the co-combustion of coal and ammonia is adopted, which includes real-time acquisition of flame images, image segmentation, feature extraction, model training and loss function evaluation, dynamic adjustment of strategy to optimize the model, and finally realizes intelligent monitoring and visualization analysis of flame combustion.
It enables intelligent, automated, and efficient monitoring of the coal and ammonia co-firing process, supporting industrial production and environmental protection, and improving the accuracy and safety of combustion status.
Smart Images
Figure CN120953685A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of coal and ammonia blending technology, and in particular to a method and apparatus for monitoring coal and ammonia blending. Background Technology
[0002] Coal-fired power generation, as a traditional energy supply method, occupies an important position in the global energy structure. However, the process of coal-fired power generation produces large amounts of carbon dioxide and other pollutants, causing serious environmental impacts. To reduce fossil fuel consumption and lower carbon emissions, countries are actively exploring low-carbon energy technologies. Ammonia, as a low-carbon fuel, does not produce carbon dioxide during combustion and is therefore considered one of the effective alternative fuels for coal-fired power generation.
[0003] Ammonia-coal co-firing technology is a technique that mixes ammonia with coal for combustion, aiming to reduce carbon emissions from coal-fired power generation and improve energy efficiency. In recent years, with in-depth research into low-carbon energy technologies, ammonia-coal co-firing technology has received widespread attention and development. This technology combines the stable supply of coal with the low-carbon characteristics of ammonia, providing a new path for achieving the low-carbon transformation of coal-fired power generation. However, pollutants and harmful gases that may be generated during the co-firing process pose threats to the environment and human health. The combustion state during coal-ammonia co-firing is dynamic, and traditional methods often struggle to monitor and accurately assess it in real time. Summary of the Invention
[0004] The purpose of this invention is to provide a method and device for monitoring the co-firing of coal and ammonia, which helps to realize the intelligent, automated and efficient co-firing process of coal and ammonia, and provides better support for industrial production and environmental protection.
[0005] To achieve the above objectives, the present invention provides a method for monitoring the co-firing of coal and ammonia, comprising the following steps:
[0006] S1. Real-time acquisition of flame images during coal-ammonia co-firing, and preprocessing of the acquired flame images;
[0007] S2. Input the preprocessed photos into the coal-ammonia co-firing monitoring model to train the coal-ammonia co-firing monitoring model, and set a loss function to evaluate the trained coal-ammonia co-firing monitoring model to obtain the evaluation results.
[0008] S3. Set up a dynamic adjustment strategy to optimize the monitoring model for coal-ammonia co-firing based on the evaluation results, and obtain the optimal monitoring model for coal-ammonia co-firing.
[0009] S4. Use the optimal coal-ammonia co-firing monitoring model to monitor the co-firing of coal and ammonia, and perform visual analysis of the co-firing to analyze the sufficiency of coal combustion.
[0010] Preferably, the preprocessing in S1 is image segmentation, specifically the following operations:
[0011] The acquired flame image is segmented, and the particles in the segmented image are counted. Particles within the boundary are directly retained. For particles at the boundary, if the area of the particle at the boundary is greater than the standard value, the particle is retained; if the area of the particle at the boundary is less than the standard value, the particle is discarded.
[0012] Preferably, the specific steps for training the model in S2 are as follows:
[0013] S2.1. Extract features from the preprocessed photos to obtain granularity features, color features, and size features. Add time-series labels to the granularity features, color features, and size features.
[0014] S2.2 Utilize the particle size characteristics with time-series labels to perform particle size analysis and obtain the particle size content in the flame at other times except for the coal feeding time;
[0015] S2.3 Utilize color and size features with time-series labels to analyze flame conditions, including the color depth and height of the flame at different heights and times;
[0016] S2.4. Based on the time sequence label, remove the particle size characteristics, color characteristics, and size characteristics of the coal that was just added.
[0017] Preferably, the detection head of the coal-ammonia co-firing monitoring model in S2 detects the flame combustion results in real time through the detection frame, including particle size characteristics, color characteristics, and size characteristics. The loss function evaluates the detected flame combustion results and outputs the loss value between the detected flame combustion results and the actual flame combustion results.
[0018] Preferably, the specific operation of the dynamic adjustment strategy in S3 is as follows: when the loss value is greater than or equal to the dynamic threshold, the analysis result is retained; when the loss value is less than a preset negative sample threshold, feature extraction needs to be reconstructed, and granular features, color features, and size features are re-extracted.
[0019] Preferably, in S4, the degree of coal combustion is analyzed based on the color depth and height of the flame at different heights and times.
[0020] This invention provides a monitoring device for coal and ammonia co-firing, applied to the aforementioned coal and ammonia co-firing monitoring method, comprising a data acquisition module, a data processing module, a model training module, and an analysis module.
[0021] The data acquisition module is used to acquire images of the flames in real time during the coal-ammonia co-firing process;
[0022] The data processing module is used to preprocess the acquired flame images;
[0023] The model training module is used to input the preprocessed photos into the coal-ammonia co-firing monitoring model to train the model, set a loss function to evaluate the trained model, obtain the evaluation structure, set a dynamic adjustment strategy to optimize the coal-ammonia co-firing monitoring model based on the evaluation results, and obtain the optimal coal-ammonia co-firing monitoring model.
[0024] The analysis module is used to monitor the co-firing of coal and ammonia using the optimal coal-ammonia co-firing detection model, and to visualize and analyze the co-firing situation.
[0025] Therefore, the coal and ammonia co-firing monitoring method and device described above in this invention helps to realize the intelligent, automated and efficient coal and ammonia co-firing process, providing better support for industrial production and environmental protection.
[0026] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description
[0027] Figure 1 This is a flowchart of a method for monitoring the co-firing of coal and ammonia according to the present invention. Detailed Implementation
[0028] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.
[0029] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.
[0030] Example 1
[0031] like Figure 1 As shown, the present invention provides a method for monitoring the co-firing of coal and ammonia, comprising the following steps:
[0032] S1. Real-time acquisition of flame images during coal-ammonia co-firing, and preprocessing of the acquired flame images;
[0033] Preprocessing involves image segmentation, specifically the following steps:
[0034] The acquired flame image is segmented, and the particles in the segmented image are counted. Particles within the boundary are directly retained. For particles at the boundary, if the area of the particle at the boundary is greater than the standard value, the particle is retained; if the area of the particle at the boundary is less than the standard value, the particle is discarded.
[0035] S2. Input the preprocessed photos into the coal-ammonia co-firing monitoring model to train the coal-ammonia co-firing monitoring model, and set a loss function to evaluate the trained coal-ammonia co-firing monitoring model to obtain the evaluation results.
[0036] The specific steps for training the model are as follows:
[0037] S2.1. Extract features from the preprocessed photos to obtain granularity features, color features, and size features. Add time-series labels to the granularity features, color features, and size features.
[0038] S2.2 Utilize the particle size characteristics with time-series labels to perform particle size analysis and obtain the particle size content in the flame at other times except for the coal feeding time;
[0039] S2.3 Utilize color and size features with time-series labels to analyze flame conditions, including the color depth and height of the flame at different heights and times;
[0040] S2.4. Based on the time series labels, remove the particle size, color, and size characteristics of the coal at the beginning of its addition to avoid the influence of the characteristic parameters at the beginning of the coal addition on the overall data and subsequent analysis.
[0041] The detection head of the coal-ammonia co-firing monitoring model detects the flame combustion results in real time through the detection frame, taking into account the particle size, color, and size characteristics. The loss function evaluates the detected flame combustion results and outputs the loss value between the detected flame combustion results and the actual flame combustion results.
[0042] S3. Set up a dynamic adjustment strategy to optimize the monitoring model for coal-ammonia co-firing based on the evaluation results, and obtain the optimal monitoring model for coal-ammonia co-firing.
[0043] The specific operation of the dynamic adjustment strategy is as follows: when the loss value is greater than or equal to the dynamic threshold, the analysis results are retained, where the dynamic threshold is a critical value range adjusted according to the real-time operating conditions; when the loss value is less than a certain preset negative sample threshold, feature extraction reconstruction is required, and granular features, color features, and size features are re-extracted, where the negative sample threshold is a critical value used to determine whether the monitoring data belongs to normal operating conditions.
[0044] S4. Use the optimal coal-ammonia co-firing monitoring model to monitor the co-firing of coal and ammonia, and perform visual analysis of the co-firing to analyze the sufficiency of coal combustion.
[0045] The color depth of the flame at different heights and times, and the height analysis, determine the completeness of coal combustion.
[0046] When the particle size is higher than the first threshold range, the flame color is dark red, and the flame height is lower than the second threshold range, it is considered incomplete combustion.
[0047] When the particle size is below the first threshold range, the flame color is blue, and the flame height is above the second threshold range, it is considered complete combustion.
[0048] When the particle size is within the first threshold range, the flame color is yellow, red, or orange, and the flame height is within the second threshold range, it is considered normal. The first threshold is the acceptable range for particle size, and the second threshold is the acceptable range for flame height.
[0049] Judging from the degree of coal combustion, it is necessary to supply more air or more coal.
[0050] This invention provides a monitoring device for coal and ammonia co-firing, applied to the aforementioned coal and ammonia co-firing monitoring method, comprising a data acquisition module, a data processing module, a model training module, and an analysis module.
[0051] The data acquisition module is used to acquire images of the flames in real time during the coal-ammonia co-firing process;
[0052] The data processing module is used to preprocess the acquired flame images;
[0053] The model training module is used to input the preprocessed photos into the coal-ammonia co-firing monitoring model to train the model, set a loss function to evaluate the trained model, obtain the evaluation structure, set a dynamic adjustment strategy to optimize the coal-ammonia co-firing monitoring model based on the evaluation results, and obtain the optimal coal-ammonia co-firing monitoring model.
[0054] The analysis module is used to monitor the co-firing of coal and ammonia using the optimal coal-ammonia co-firing detection model, and to visualize and analyze the co-firing situation.
[0055] Therefore, the coal and ammonia co-firing monitoring method and device described above in this invention helps to realize the intelligent, automated and efficient coal and ammonia co-firing process, providing better support for industrial production and environmental protection.
[0056] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.
Claims
1. A method for monitoring the co-firing of coal and ammonia, characterized in that: Includes the following steps: S1. Real-time acquisition of flame images during coal-ammonia co-firing, and preprocessing of the acquired flame images; S2. Input the preprocessed photos into the coal-ammonia co-firing monitoring model to train the coal-ammonia co-firing monitoring model, and set a loss function to evaluate the trained coal-ammonia co-firing monitoring model to obtain the evaluation results. S3. Set up a dynamic adjustment strategy to optimize the monitoring model for coal-ammonia co-firing based on the evaluation results, and obtain the optimal monitoring model for coal-ammonia co-firing. S4. Use the optimal coal-ammonia co-firing monitoring model to monitor the co-firing of coal and ammonia, and perform visual analysis of the co-firing to analyze the sufficiency of coal combustion.
2. The method for monitoring the co-firing of coal and ammonia according to claim 1, characterized in that: The preprocessing in S1 is image segmentation, specifically the following operations: The acquired flame image is segmented, and the particles in the segmented image are counted. Particles within the boundary are directly retained. For particles at the boundary, if the area of the particle at the boundary is greater than the standard value, the particle is retained; if the area of the particle at the boundary is less than the standard value, the particle is discarded.
3. The method for monitoring the co-firing of coal and ammonia according to claim 2, characterized in that: The specific steps for training the model in S2 are as follows: S2.
1. Extract features from the preprocessed photos to obtain granularity features, color features, and size features. Add time-series labels to the granularity features, color features, and size features. S2.2 Utilize the particle size characteristics with time-series labels to perform particle size analysis and obtain the particle size content in the flame at other times except for the coal feeding time; S2.3 Utilize color and size features with time-series labels to analyze flame conditions, including the color depth and height of the flame at different heights and times; S2.
4. Based on the time sequence label, remove the particle size characteristics, color characteristics, and size characteristics of the coal that was just added.
4. The method for monitoring the co-firing of coal and ammonia according to claim 3, characterized in that: The detection head of the coal-ammonia co-combustion monitoring model in S2 detects the flame combustion results in real time through the detection frame, including particle size, color, and size characteristics. The loss function evaluates the detected flame combustion results and outputs the loss value between the detected flame combustion results and the actual flame combustion results.
5. The method for monitoring the co-firing of coal and ammonia according to claim 4, characterized in that: The specific operation of the dynamic adjustment strategy in S3 is as follows: when the loss value is greater than or equal to the dynamic threshold, the analysis results are retained; when the loss value is less than a preset negative sample threshold, feature extraction needs to be reconstructed, and granular features, color features, and size features are re-extracted.
6. The method for monitoring the co-firing of coal and ammonia according to claim 5, characterized in that: S4 analyzes the completeness of coal combustion based on the color depth and height of the flame at different heights and times.
7. A device for monitoring the co-firing of coal and ammonia, applied to the method for monitoring the co-firing of coal and ammonia as described in claims 1-6, characterized in that: It includes a data acquisition module, a data processing module, a model training module, and an analysis module. The data acquisition module is used to acquire images of the flames in real time during the coal-ammonia co-firing process; The data processing module is used to preprocess the acquired flame images; The model training module is used to input the preprocessed photos into the coal-ammonia co-firing monitoring model to train the model, set a loss function to evaluate the trained model, obtain the evaluation structure, set a dynamic adjustment strategy to optimize the coal-ammonia co-firing monitoring model based on the evaluation results, and obtain the optimal coal-ammonia co-firing monitoring model. The analysis module is used to monitor the co-firing of coal and ammonia using the optimal coal-ammonia co-firing detection model, and to visualize and analyze the co-firing situation.