Intelligent agent spraying decoking system for energy conservation and decoking of hearth

Abstract

The invention discloses an intelligent agent spraying decoking system for energy-saving decoking of a hearth, and relates to the technical field of decoking of industrial boilers, the intelligent agent spraying decoking system comprises a sensing module used for collecting coke scale characteristic data, and the data comprises the temperature, thickness, component distribution and adhesion strength of coke scale; the module is combined with a temperature sensor, a component analyzer and image detection equipment for cooperative work, and multi-dimensional scorch scale characteristic information is obtained in real time. According to the intelligent agent spraying decoking system for hearth energy saving and decoking, through cooperative use of the sensing module, the control module, the agent storage module, the agent spraying module and the feedback module, intelligent and accurate decoking is achieved, the diversity of coke scales can be effectively dealt with, the heat efficiency of a boiler is improved, the energy cost is reduced, and the safety problem caused by the coke scales can be prevented; the problems that a traditional decoking method is low in efficiency, poor in applicability and insufficient in adaptability to coke scale types are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial boiler descaling, and particularly to an intelligent chemical spraying descaling system for energy-saving descaling in a furnace. Background Art

[0002] During the operation of industrial boilers, due to the complex fuel composition and variable combustion conditions, various types of coke deposits are likely to form on the inner wall of the furnace and the heat exchange surface; these coke deposits not only reduce the thermal efficiency of the boiler but may also cause safety problems; currently, the descaling methods on the market are mostly manual cleaning or fixed chemical spraying, and these methods have problems such as low efficiency, poor applicability, and insufficient adaptability to the types of coke deposits; therefore, there is an urgent need for an intelligent and efficient descaling system that can target different coke deposit characteristics to improve the boiler operation efficiency and reduce the operation cost. Summary of the Invention

[0003] To solve the above technical problems, the present invention is achieved through the following technical solutions: An intelligent chemical spraying descaling system for energy-saving descaling in a furnace, comprising:

[0004] A sensing module for collecting coke deposit characteristic data, where the data includes the temperature, thickness, composition distribution, and adhesion strength of the coke deposit; this module works in collaboration with a temperature sensor, a composition analyzer, and an image detection device to obtain multi-dimensional coke deposit characteristic information in real time;

[0005] A control module for analyzing the coke deposit characteristic data and generating a spraying strategy, including: using a built-in deep learning algorithm to classify and extract features from the coke deposit characteristic data, identifying the coke deposit type according to the model, combining the chemical agent database and the coke deposit type, automatically selecting the optimal chemical agent formula, obtaining a chemical agent combination, and generating specific spraying parameters, generating a spraying strategy based on the chemical agent combination and the spraying parameters; at the same time, adjusting the spraying strategy in combination with the descaling quality of the feedback module to obtain an optimized spraying plan, forming a closed-loop optimization mechanism; and generating a control instruction based on the optimized spraying plan and sending the control instruction to the spraying module to perform the chemical agent spraying operation;

[0006] A chemical agent storage module for storing descaling chemicals with different compositions, the chemical agent storage module includes several stainless-steel-made storage tanks to ensure that the storage tanks are not damaged during long-term storage of corrosive chemicals; the storage tanks are equipped with pressure relief valves, temperature regulation, and humidity control devices to prevent excessive internal pressure, ensure the stability of the chemicals, ensure a stable chemical agent storage environment, and avoid the influence of too high temperature or too low humidity on the chemical agent performance; used for classifying and storing oxidants, alkaline neutralizers, and multi-component mixed chemicals in separate tanks to avoid mixing reactions between different chemicals;

[0007] A spraying module for implementing chemical agent spraying, the spraying module includes an adjustable nozzle for adjusting the spraying angle and pressure;

[0008] A feedback module for monitoring the decoking quality and optimizing the spraying strategy.

[0009] Preferably, the sensing module combines a temperature sensor, a composition analyzer, and an image detection device to work together to obtain multi-dimensional coke scale characteristic information in real time. The process is as follows: First, measure the temperature gradient of the coke scale surface and the surrounding environment through the temperature sensor; then perform spectral detection through the composition analyzer to analyze the main chemical components of the coke scale; finally, combine the high-definition camera and the image processing algorithm through the image detection device to generate a coke scale distribution map and thickness measurement data; based on the temperature gradient, main chemical components, coke scale distribution map, and thickness measurement data, jointly constitute multi-dimensional coke scale characteristic information, and transmit the coke scale characteristic information to the control module in real time through the Internet of Things technology; among them, the temperature sensor measures the temperature gradient of the coke scale surface and the surrounding environment, uses infrared thermometry technology to perform non-contact high-precision temperature measurement on the coke scale surface, and the infrared sensor works in the wavelength range of 3-14 microns to obtain the coke scale surface temperature. At the same time, combine the K-type thermocouple sensor to sample the furnace environment temperature in real time to obtain the furnace environment temperature; based on the coke scale surface temperature and the furnace environment temperature, generate a temperature gradient through multi-point calibration and filtering algorithm processing.

[0010] Preferably, the composition analyzer analyzes the main chemical components of the coke scale through spectral detection: carbon, sulfur, calcium, silicon; at the same time, further identifies complex organic and inorganic compounds through Fourier transform infrared spectroscopy (FTIR) technology to provide comprehensive coke scale composition characteristic data; among them, the process of analyzing the main chemical components of the coke scale by the composition analyzer through spectral detection includes:

[0011] Use X-ray fluorescence spectroscopy (XRF) technology to perform qualitative and quantitative analysis of the elemental composition in the coke scale sample. Set the main chemical components: carbon, sulfur, calcium, silicon as the detection targets, including: using a mechanical or automatic sampling device to extract a trace sample from the coke scale surface and prepare it into powder or sheet to make a coke scale sample; then load a standard sample on the XRF device to calibrate the energy and sensitivity of the detector to ensure the analysis accuracy; place the prepared coke scale sample in the sample chamber of the XRF device, generate secondary fluorescence through high-energy X-ray irradiation, capture the fluorescence signal by a high-sensitivity detector, and convert it into a digital signal, record the intensity and energy of the characteristic peaks of each element; then use built-in or external analysis software to match the fluorescence intensity with the energy wavelength database, calculate the types and contents of elements in the coke scale sample; generate a qualitative and quantitative report of the coke scale composition;

[0012] The process of further identifying complex organic and inorganic compounds through FTIR (Fourier Transform Infrared Spectroscopy) technology is as follows: The coke deposit sample is crushed and prepared into thin slices or mixed with a solvent to form a liquid sample to suit FTIR analysis; Select a wavenumber range of 400 - 4000 cm - -1 in the FTIR equipment and calibrate the interferometer; Then irradiate the sample with an infrared light source, record its absorption spectrum, and obtain infrared absorption peaks through the interferometer; Match the spectral data with the NIST library to identify the organic functional groups and inorganic compound characteristics in the sample, generate a detailed spectral analysis report of the coke deposit composition, provide support for optimizing the coke removal plan, and provide comprehensive coke deposit composition characteristic data.

[0013] Preferably, through an image detection device combining a high-definition camera and an image processing algorithm, a coke deposit distribution map and thickness measurement data are generated. The process includes: Obtain high-resolution images of the coke deposit area in the furnace through a high-definition camera, and use multi-angle shooting technology to ensure comprehensive image coverage, including: Install high-definition cameras at various corner positions in the furnace for multi-view shooting to capture images of the coke deposit area from different angles and ensure that the acquired image data comprehensively covers the coke deposit distribution area; And use a mechanical control arm or a movable camera to dynamically adjust the shooting angle to adapt to the complex spatial structure and coke deposit distribution changes in the furnace; Subsequently, perform time synchronization processing on the images taken by the multi-angle cameras for subsequent image stitching and analysis to obtain the original images; Denoise, enhance the contrast, and detect the edges of the original images to highlight the coke deposit features; Then use a convolutional neural network to automatically identify the coke deposit distribution area in the images, generate a binary image, and convert the obtained data into a standardized format to obtain the coke deposit distribution map. Combine the gray value distribution of the images and known calibration parameters to calculate the relative thickness distribution of the coke deposit, and convert the obtained data into a standardized format to obtain the thickness measurement data.

[0014] Preferably, an in-built deep learning algorithm is used to classify and extract features from the fouling characteristic data, and the fouling type is identified according to the model; it includes: preprocessing the input data to remove noise and outliers to ensure data quality, where the input data includes temperature gradient, main chemical components, and thickness measurement data; constructing a deep learning model based on historical data, including using a convolutional neural network (CNN) to extract features from image data and combining with a multi-layer perceptron (MLP) to process compositional analysis data; including: collecting and annotating a large number of fouling sample data, including image data and compositional analysis data, ensuring that the data covers a variety of fouling types and complex characteristics, preprocessing the image data, including normalization, data augmentation, rotation, and scaling, to improve the robustness of the model to different samples, constructing a convolutional neural network (CNN) to extract the spatial features of the fouling image data, and at the same time designing a multi-layer perceptron (MLP) to process numerical compositional analysis data, and the two models achieve joint learning through a feature fusion module, using the annotated data for supervised learning, adopting the backpropagation algorithm to optimize the network weights, and introducing an early stopping mechanism to prevent overfitting, evaluating the classification accuracy and feature extraction ability of the model through a validation set, adjusting the network structure or parameters to optimize the performance, saving the finally trained model in a deployable format for real-time fouling type identification and feature analysis; subsequently, the preprocessed fouling characteristic data is input into the trained model, and the fouling classification result and key features are output to obtain the fouling type.

[0015] Preferably, in combination with the chemical agent database and the fouling type, the optimal chemical agent formula is automatically selected to obtain a chemical agent combination, and specific spraying parameters are generated. The process of obtaining the spraying strategy is as follows: First, extract the optimal chemical agent formula matching the fouling type from the chemical agent database, and select the chemical agent combination according to the composition, thickness, and distribution characteristics of the fouling; then, according to the adhesion strength and distribution range of the fouling, use the built-in algorithm to calculate the required spraying parameters, where the spraying parameters include spraying pressure, chemical agent dosage, spraying angle range, spraying path, and chemical agent formula; generate the spraying strategy according to the chemical agent combination and the spraying parameters.

[0016] Preferably, in combination with the coke removal quality of the feedback module, the spraying strategy is adjusted to obtain an optimized spraying plan, and a closed-loop optimization mechanism is formed, including: First, the feedback module monitors the residual coke scale, the quality of chemical agent spraying, and the boiler operation status data in real time after coke removal to obtain the coke removal quality. Then, through the built-in evaluation model of the control module, it analyzes whether the current coke removal quality meets the expected set standards, and the set standards include the expected coke scale removal rate and the boiler thermal efficiency. If the coke removal quality does not meet the set standards, the spraying parameters are analyzed, and the spraying parameters in the spraying strategy are adjusted to ensure uniform chemical agent coverage and maximum efficiency, generate an optimized spraying plan, and generate a control command at the same time. And the control command is sent to the spraying module in real time for adjusting the pressure, angle or dosage of chemical agent spraying in real time. At the same time, the quality data of the optimized adjustment after each execution of the optimized spraying plan is recorded, and the quality data is used to train the closed-loop optimization mechanism of the control module to improve the intelligent level of subsequent spraying strategies, and finally form a closed-loop optimization mechanism.

[0017] Preferably, the chemical agent storage module further includes: By equipping the storage tank with a liquid level sensor, a concentration sensor and a temperature sensor, it monitors the remaining amount, concentration and storage conditions of the chemical agent in the storage tank in real time, generates storage tank status data, and transmits the storage tank status data to the control module through a wireless transmission module. The control module generates an assistance command based on the storage tank status data, and the assistance command is used to generate a coordination command in combination with the control command, and controls the spraying module to spray the chemical agent based on the coordination command.

[0018] Preferably, the storage tank is connected to the spraying module through a corrosion-resistant pipeline and is equipped with a chemical agent pump, and the chemical agent pump is used to transport the chemical agent according to the assistance command in the coordination command of the control module to ensure no leakage and high transportation efficiency.

[0019] The present invention provides an intelligent chemical agent spraying and coke removal system for furnace energy-saving and coke removal, which has the following beneficial effects:

[0020] This intelligent chemical agent spraying and coke removal system for furnace energy-saving and coke removal realizes intelligent and precise coke removal through the collaborative use of sensing, control, chemical agent storage, spraying and feedback modules, can effectively cope with the diversity of coke scale, improve the boiler thermal efficiency, reduce the energy cost, and can also prevent safety problems caused by coke scale, solving the problems of low efficiency, poor applicability and insufficient adaptability to coke scale types of traditional coke removal methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a framework schematic diagram of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0022] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. The embodiments of the present invention are given for the purpose of illustration and description, and are not exhaustive or limit the present invention to the disclosed form. Many modifications and variations are obvious to those of ordinary skill in the art. The embodiments are selected and described to better illustrate the principles and practical applications of the present invention, and enable those of ordinary skill in the art to understand the present invention and design various embodiments with various modifications suitable for specific purposes.

[0023] As Figure 1 shown, the present invention provides a technical solution: an intelligent chemical agent spraying and coke removal system for furnace energy saving and coke cleaning, comprising:

[0024] A sensing module for collecting coke scale characteristic data, where the data includes the temperature, thickness, composition distribution, and adhesion strength of the coke scale; this module works in cooperation with a temperature sensor, a composition analyzer, and an image detection device to obtain multi-dimensional coke scale characteristic information in real time;

[0025] A control module for analyzing the coke scale characteristic data and generating a spraying strategy, including: using a built-in deep learning algorithm to classify and extract features from the coke scale characteristic data, identifying the coke scale type according to the model, combining the chemical agent database and the coke scale type, automatically selecting the optimal chemical agent formula, obtaining a chemical agent combination, and generating specific spraying parameters, generating a spraying strategy based on the chemical agent combination and the spraying parameters; at the same time, adjusting the spraying strategy in combination with the coke cleaning quality of the feedback module to obtain an optimized spraying plan, and forming a closed-loop optimization mechanism; and generating a control instruction based on the optimized spraying plan and sending the control instruction to the spraying module to perform the chemical agent spraying operation;

[0026] A chemical agent storage module for storing coke removal chemical agents with different compositions. The chemical agent storage module includes several stainless steel-made storage tanks to ensure that the storage tanks are not damaged during long-term storage of corrosive chemical agents; the storage tanks are equipped with pressure relief valves, temperature regulation, and humidity control devices to prevent excessive internal pressure, ensure the stability of the chemical agents, ensure a stable storage environment for the chemical agents, and avoid the influence of too high temperature or too low humidity on the performance of the chemical agents; for classifying and storing oxidants, alkaline neutralizers, and multi-component mixed chemical agents in separate tanks to avoid mixing reactions between different chemical agents;

[0027] A spraying module for implementing chemical agent spraying. The spraying module includes adjustable nozzles for adjusting the spraying angle and pressure;

[0028] A feedback module for monitoring the coke cleaning quality and optimizing the spraying strategy.

[0029] By means of multi - means collaboration of the sensing module to collect data, comprehensive fouling information is provided, laying a foundation for precise descaling, reducing blindness; through the intelligent decision - making of the deep - learning algorithm in the control module, the pertinence and efficiency of descaling are improved, and the closed - loop optimization adapts to changes, enhancing the quality and stability of descaling; through the classified storage and environmental control of the chemical agent storage module, the quality and dispensing flexibility of chemical agents are guaranteed, facilitating long - term stable operation; through the adjustable nozzles of the spraying module, different working conditions are adapted to ensure uniform coverage of chemical agents and enhance the descaling effect; through the feedback module to construct an optimization cycle, the system continuously self - improves, ensuring the efficient, stable and energy - saving operation of the boiler.

[0030] Through the collaborative operation of a temperature sensor, a composition analyzer and an image detection device, the sensing module can obtain fouling characteristic information in real - time, accurately and multi - dimensionally, including temperature gradient, chemical composition, distribution and thickness, etc. This enables the system to comprehensively and deeply understand the fouling situation, just like drawing a detailed portrait of the fouling, thus providing a very accurate basis for subsequent descaling operations, avoiding the blind operation caused by incomplete information in traditional descaling methods, and greatly improving the pertinence and effectiveness of descaling.

[0031] The sensing module combines a temperature sensor, a composition analyzer and an image detection device to work together to obtain multi - dimensional fouling characteristic information in real - time. The process is as follows: First, the temperature sensor measures the temperature gradient between the fouling surface and the surrounding environment; then, the composition analyzer conducts spectral detection to analyze the main chemical components of the fouling; finally, the image detection device combines a high - definition camera and an image - processing algorithm to generate a fouling distribution map and thickness measurement data. Based on the temperature gradient, main chemical components, fouling distribution map and thickness measurement data, multi - dimensional fouling characteristic information is formed and transmitted to the control module in real - time through Internet of Things technology. Among them, the temperature sensor measures the temperature gradient between the fouling surface and the surrounding environment, uses infrared thermometry technology to conduct non - contact high - precision temperature measurement on the fouling surface, and the infrared sensor works in the wavelength range of 3 - 14 microns to obtain the fouling surface temperature. At the same time, the K - type thermocouple sensor is combined to sample the furnace environment temperature in real - time to obtain the furnace environment temperature. Based on the fouling surface temperature and the furnace environment temperature, through multi - point calibration and filtering algorithm processing, the temperature gradient is generated.

[0032] The composition analyzer analyzes the main chemical components of the fouling through spectral detection: carbon, sulfur, calcium, silicon; at the same time, the infrared spectroscopy FTIR technology is used to further identify complex organic and inorganic compounds to provide comprehensive fouling composition characteristic data. Among them, the process of the composition analyzer analyzing the main chemical components of the fouling through spectral detection includes:

[0033] Qualitative and quantitative analysis of the elemental composition of coke scale samples is carried out using X-ray fluorescence spectroscopy (XRF). The main chemical components, namely carbon, sulfur, calcium, and silicon, are set as the detection targets, including: extracting a trace sample from the coke scale surface using a mechanical or automatic sampling device and preparing it into a powder or flake to form a coke scale sample; subsequently, loading a standard sample on the XRF device to calibrate the energy and sensitivity of the detector to ensure analysis accuracy; placing the prepared coke scale sample in the sample chamber of the XRF device, generating secondary fluorescence through irradiation with high-energy X-rays, capturing the fluorescence signal by a high-sensitivity detector, and converting it into a digital signal, recording the intensity and energy of the characteristic peaks of each element; then using built-in or external analysis software to match the fluorescence intensity with the energy wavelength database, calculating the types and contents of elements in the coke scale sample; generating a qualitative and quantitative report on the coke scale composition;

[0034] The process of further identifying complex organic and inorganic compounds through Fourier transform infrared spectroscopy (FTIR) is as follows: crushing the coke scale sample and preparing it into a thin film or mixing it with a solvent to form a liquid sample to suit FTIR analysis; selecting a wavenumber range of 400 - 4000 cm - -1 in the FTIR device and calibrating the interferometer; then irradiating the sample with an infrared light source, recording its absorption spectrum, and obtaining infrared absorption peaks through the interferometer; matching the spectral data with the NIST library to identify the organic functional groups and inorganic compound characteristics in the sample, generating a detailed spectral analysis report on the coke scale composition, providing support for optimizing the coke removal plan to provide comprehensive coke scale composition characteristic data.

[0035] Combined measurement by temperature sensors accurately obtains the temperature gradient, providing key thermal basis for the coke removal strategy to ensure safety and effectiveness; while through the multi-technique combination of the composition analyzer, deeply analyzing the chemical composition of the coke scale, helping to accurately select drugs, reducing costs and pollution risks; finally, through multi-angle imaging and intelligent processing of the imaging detection device, visually presenting the distribution and thickness of the coke scale, improving the accuracy and visualization of coke removal.

[0036] It should be further noted that in the specific implementation process, the control module, based on the deep learning algorithm, conducts intelligent classification and feature extraction on the collected coke scale characteristic data, and then automatically matches the optimal drug formula from the drug database and generates a detailed and accurate spraying strategy; this intelligent decision-making process ensures a high degree of compatibility between the drug and the coke scale, just like tailoring a coke removal plan for each type of coke scale, enabling the drug to fully exert its coke removal efficiency, effectively reducing drug waste, improving the coke removal efficiency, shortening the coke removal operation time, and at the same time reducing the risk of damage to the furnace equipment that may be caused by improper use of the drug.

[0037] Through an imaging detection device combined with a high-definition camera and image processing algorithms, a fouling distribution map and thickness measurement data are generated. The process includes: obtaining high-resolution images of the fouling area in the furnace through a high-definition camera, and using multi-angle shooting technology to ensure comprehensive image coverage, including: installing high-definition cameras at various corner positions in the furnace for multi-view shooting, capturing images of the fouling area from different angles to ensure that the acquired image data comprehensively covers the fouling distribution area; and using a mechanical control arm or a movable camera to dynamically adjust the shooting angle to adapt to the complex spatial structure and fouling distribution changes in the furnace; subsequently, performing time synchronization processing on the images taken by the multi-angle cameras to facilitate subsequent image stitching and analysis, obtaining the original images; denoising, enhancing the contrast, and detecting the edges of the original images to highlight the fouling features; then using a convolutional neural network to automatically identify the fouling distribution area in the images, generating a binary image, and converting the obtained data into a standardized format to obtain the fouling distribution map, and combining the gray value distribution of the images and known calibration parameters to calculate the relative thickness distribution of the fouling, and converting the obtained data into a standardized format to obtain the thickness measurement data.

[0038] Using the built-in deep learning algorithm to classify and extract fouling characteristic data, and identifying the fouling type according to the model; including: preprocessing the input data to remove noise and outliers to ensure data quality, and the input data includes temperature gradient, main chemical components, and thickness measurement data; constructing a deep learning model based on historical data, including using a convolutional neural network (CNN) to extract features from the imaging data, and combining a multi-layer perceptron (MLP) to process the composition analysis data; including: collecting and annotating a large number of fouling sample data, including imaging data and composition analysis data, ensuring that the data covers a variety of fouling types and complex characteristics, preprocessing the imaging data, including normalization, data augmentation, rotation, and scaling, to improve the robustness of the model to different samples, constructing a convolutional neural network (CNN) to extract the spatial features of the fouling imaging data, and at the same time designing a multi-layer perceptron (MLP) to process the numerical composition analysis data, and the two models achieve joint learning through a feature fusion module, using the annotated data for supervised learning, adopting the backpropagation algorithm to optimize the network weights, and introducing an early stopping mechanism to prevent overfitting, evaluating the classification accuracy and feature extraction ability of the model through a validation set, adjusting the network structure or parameters to optimize the performance, and saving the finally trained model in a deployable format for real-time fouling type identification and feature analysis; subsequently, inputting the preprocessed fouling characteristic data into the trained model, and outputting the fouling classification result and key features to obtain the fouling type.

[0039] Combined with the chemical agent database and the type of coke fouling, automatically select the optimal chemical agent formulation, obtain the chemical agent combination, and generate specific spraying parameters. The process of generating the spraying strategy based on the chemical agent combination and spraying parameters is as follows: First, extract the optimal chemical agent formulation that matches the type of coke fouling from the chemical agent database, and select the chemical agent combination according to the composition, thickness, and distribution characteristics of the coke fouling. Then, according to the adhesion strength and distribution range of the coke fouling, use the built-in algorithm to calculate the required spraying parameters. The spraying parameters include spraying pressure, chemical agent dosage, spraying angle range, spraying path, and chemical agent formulation. Generate the spraying strategy based on the chemical agent combination and spraying parameters.

[0040] Combine the coke cleaning quality of the feedback module to adjust the spraying strategy, obtain the optimized spraying plan, and form a closed-loop optimization mechanism at the same time. It includes: First, the feedback module monitors the residual coke fouling situation, chemical agent spraying quality, and boiler operation status data after coke cleaning in real time to obtain the coke cleaning quality. Then, through the built-in evaluation model of the control module, analyze whether the current coke cleaning quality meets the expected set standards. The set standards include the expected coke fouling removal rate and boiler thermal efficiency. If the coke cleaning quality does not meet the set standards, analyze the spraying parameters and adjust the spraying parameters in the spraying strategy to ensure uniform chemical agent coverage and maximum efficiency, generate the optimized spraying plan, and generate control instructions at the same time. And send the control instructions to the spraying module in real time for adjusting the pressure, angle, or dosage of chemical agent spraying in real time. At the same time, record the quality data of the optimized adjustment after each execution of the optimized spraying plan. The quality data is used to train the closed-loop optimization mechanism of the control module to improve the intelligent level of subsequent spraying strategies, and finally form a closed-loop optimization mechanism.

[0041] The chemical agent storage module also includes: By equipping the storage tank with a liquid level sensor, a concentration sensor, and a temperature sensor, monitor the remaining amount, concentration, and storage conditions of the chemical agent in the storage tank in real time, generate the storage tank status data, and transmit the storage tank status data to the control module through the wireless transmission module. The control module generates an assistance instruction based on the storage tank status data, and the assistance instruction is used to generate a coordination instruction in combination with the control instruction, and control the spraying module to spray the chemical agent based on the coordination instruction. Among them, the monitoring and data transmission of the storage tank sensor facilitate mastering the chemical agent storage status and planning in advance to ensure the continuity of operations. And through the coordination instruction of the control module based on the storage tank data, ensure the stable and reliable supply of chemical agents and reduce the impact of problems in the supply link.

[0042] The storage tank is connected to the spraying module through a corrosion-resistant pipeline and is equipped with a chemical agent pump. The chemical agent pump is used to transport the chemical agent according to the assistance instruction in the coordination instruction of the control module. The corrosion-resistant pipeline cooperates with the chemical agent pump to ensure the safe and stable transportation of the chemical agent, achieve precise spraying, and improve the consistency and reliability of coke removal.

[0043] Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those of ordinary skill in the art and related fields based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention. Structures, devices, and operation methods not specifically described and explained in the present invention shall be implemented by conventional means in the art without special instructions and limitations.

Claims

1. An intelligent agent spraying decoking system for energy-saving and decoking furnace, characterized in that: include: A sensor module for collecting coke scale characteristic data, wherein the data includes the temperature, thickness, composition distribution and adhesion strength of the coke scale; The module works in conjunction with temperature sensors, component analyzers and image detection equipment to obtain multi-dimensional coke fouling characteristic information in real time; A control module is used to analyze the coke scale characteristic data and generate a spraying strategy, including: using a built-in deep learning algorithm to classify and extract features from the coke scale characteristic data, identifying the coke scale type according to the model, automatically selecting the optimal reagent formula in combination with the reagent database and the coke scale type, obtaining a chemical reagent combination, and generating specific spraying parameters, and generating a spraying strategy based on the chemical reagent combination and the spraying parameters; at the same time, adjusting the spraying strategy in combination with the coke cleaning quality of the feedback module to obtain an optimized spraying scheme, and forming a closed-loop optimization mechanism; and generating a control instruction based on the optimized spraying scheme, and sending the control instruction to the spraying module to execute the reagent spraying operation; The reagent storage module is used to store decoking reagents of different components. The reagent storage module includes a plurality of storage tanks made of stainless steel. The storage tanks are equipped with pressure release valves, temperature adjustment and humidity control devices. The storage tanks are used to store oxidants, alkaline neutralizers and multi-component mixed reagents in classified tanks. A spraying module, used for spraying the agent, the spraying module comprising an adjustable nozzle for adjusting the spraying angle and pressure; Feedback module to monitor defocusing quality and optimize application strategy.

2. The intelligent agent spraying decoking system for energy-saving decoking in the furnace according to claim 1 is characterized in that: The sensing module works in conjunction with a temperature sensor, a component analyzer and an image detection device to obtain multi-dimensional coke scale characteristic information in real time, and the process is as follows: first, the temperature gradient of the coke scale surface and the surrounding environment is measured by a temperature sensor; then, the main chemical components of the coke scale are analyzed by a component analyzer through spectral detection; finally, a coke scale distribution map and thickness measurement data are generated by an image detection device combined with a high-definition camera and an image processing algorithm; multi-dimensional coke scale characteristic information is jointly formed based on the temperature gradient, the main chemical components, the coke scale distribution map and the thickness measurement data, and the coke scale characteristic information is transmitted to the control module in real time through the Internet of Things technology; wherein, the temperature sensor measures the temperature gradient of the coke scale surface and the surrounding environment, and uses infrared temperature measurement technology to perform non-contact high-precision temperature measurement on the coke scale surface, and the infrared sensor works within a wavelength range of 3-14 microns to obtain the coke scale surface temperature, and at the same time, a K-type thermocouple sensor is combined to sample the furnace environment temperature in real time to obtain the furnace environment temperature; based on the coke scale surface temperature and the furnace environment temperature, a temperature gradient is generated through multi-point calibration and filtering algorithm processing.

3. The intelligent agent spraying decoking system for energy-saving decoking in the furnace according to claim 2 is characterized in that: The component analyzer analyzes the main chemical components of coke scale by spectral detection: carbon, sulfur, calcium, silicon; and further identifies complex organic matter and inorganic compounds by infrared spectroscopy FTIR technology; wherein the process of the component analyzer analyzing the main chemical components of coke scale by spectral detection includes: X-ray fluorescence spectroscopy (XRF) technology is used to conduct qualitative and quantitative analysis of the elemental composition in the coke scale sample, and the main chemical components: carbon, sulfur, calcium, and silicon are set as detection targets, including: using a mechanical or automatic sampling device to extract a trace sample from the coke scale surface, and preparing it into powder or flakes to make a coke scale sample; then loading a standard sample on the XRF device to calibrate the energy and sensitivity of the detector; placing the prepared coke scale sample in the sample compartment of the XRF device, generating secondary fluorescence through high-energy X-ray irradiation, and capturing the fluorescence signal with a high-sensitivity detector and converting it into a digital signal, recording the intensity and energy of the characteristic peaks of each element; then using built-in or external analysis software, matching the fluorescence intensity with the energy wavelength database to calculate the type and content of the elements in the coke scale sample; generating a qualitative and quantitative report on the coke scale composition; The process of further identifying complex organic and inorganic compounds by infrared spectroscopy FTIR technology is as follows: crush the coke sample and prepare it into thin slices or mix it with a solvent to form a liquid sample; select 400-4000cm in the FTIR device - 1 wavenumber range and calibrate the interferometer; then use an infrared light source to irradiate the sample, record its absorption spectrum, and obtain the infrared absorption peak through the interferometer; match the spectral data with the NIST library, identify the organic functional groups and inorganic compound characteristics in the sample, and generate a detailed spectral analysis report of the coke composition.

4. The intelligent agent spraying decoking system for energy-saving decoking in furnace according to claim 3 is characterized in that: The coke scale distribution map and thickness measurement data are generated by combining the image detection equipment with a high-definition camera and an image processing algorithm. The process includes: obtaining a high-resolution image of the coke scale area in the furnace through a high-definition camera, and using multi-angle shooting technology to ensure comprehensive image coverage, including: installing high-definition cameras in various corners of the furnace to perform multi-angle shooting, capture images of the coke scale area at different angles, and ensure that the acquired image data fully covers the coke scale distribution area; and using a mechanical control arm or a movable camera to dynamically adjust the shooting angle; then performing time synchronization processing on the images taken by the multi-angle camera to obtain the original image; denoising, contrast enhancement and edge detection on the original image; then using a convolutional neural network to automatically identify the coke scale distribution area in the image to generate a binary image, and converting the obtained data into a standardized format to obtain a coke scale distribution map, and combining the gray value distribution of the image and known calibration parameters to calculate the relative thickness distribution of the coke scale, and converting the obtained data into a standardized format to obtain thickness measurement data.

5. The intelligent agent spraying decoking system for energy-saving decoking in the furnace according to claim 4 is characterized in that: Use built-in deep learning algorithms to classify and extract features from coke fouling characteristic data, and identify coke fouling types based on the model; including: preprocessing input data to remove noise and outliers, input data includes temperature gradient, main chemical composition, thickness measurement data; building a deep learning model based on historical data, including using convolutional neural network CNN to extract features from image data, and combining multi-layer perceptron MLP to process component analysis data; including: collecting and annotating coke fouling sample data, including image data and component analysis data, preprocessing image data, including normalization, data enhancement, rotation, scaling, building a convolutional neural network CNN for The spatial features of the coke scale image data are extracted, and a multi-layer perceptron (MLP) is designed to process the numerical component analysis data. The two models realize joint learning through the feature fusion module, and the labeled data are used for supervised learning. The back propagation algorithm is used to optimize the network weights, and an early stopping mechanism is introduced to prevent overfitting. The classification accuracy and feature extraction ability of the model are evaluated through the validation set, and the network structure or parameters are adjusted to optimize the performance. The final trained model is saved in a deployable format for real-time coke scale type identification and feature analysis. Subsequently, the preprocessed coke scale characteristic data is input into the trained model, and the coke scale classification results and key features are output to obtain the coke scale type.

6. The intelligent agent spraying decoking system for energy-saving decoking in the furnace according to claim 5 is characterized in that: Combined with the reagent database and coke scale type, the optimal reagent formula is automatically selected to obtain the chemical reagent combination, and specific spraying parameters are generated. The spraying strategy is generated based on the chemical reagent combination and spraying parameters. The process of obtaining the spraying strategy is as follows: first, the optimal reagent formula matching the coke scale type is extracted from the reagent database, and the chemical reagent combination is selected according to the composition, thickness and distribution characteristics of the coke scale; then, according to the adhesion strength and distribution range of the coke scale, the required spraying parameters are calculated using the built-in algorithm. The spraying parameters include spraying pressure, reagent dosage and injection angle range, spraying path, and reagent formula; the spraying strategy is generated based on the chemical reagent combination and spraying parameters.

7. The intelligent agent spraying decoking system for energy-saving decoking in the furnace according to claim 6 is characterized in that: Combined with the decoking quality of the feedback module, the spraying strategy is adjusted to obtain an optimized spraying plan, and a closed-loop optimization mechanism is formed at the same time; including: first, the feedback module monitors the coke scale residue after decoking, the quality of the agent spraying and the boiler operation status data in real time to obtain the decoking quality; then, through the built-in evaluation model of the control module, it is analyzed whether the current decoking quality meets the expected set standards, and the set standards include the expected coke scale removal rate and boiler thermal efficiency; if the decoking quality does not meet the set standards, the spraying parameters are analyzed, and the spraying parameters in the spraying strategy are adjusted to generate an optimized spraying plan, and control instructions are generated at the same time; and the control instructions are sent to the spraying module in real time for real-time adjustment of the pressure, angle or dosage of the agent spraying; at the same time, the quality data of the optimization adjustment after each execution of the optimized spraying plan is recorded.

8. The intelligent agent spraying decoking system for energy-saving decoking in furnace according to claim 7 is characterized in that: The agent storage module also includes: by equipping the storage tank with a liquid level sensor, a concentration sensor and a temperature sensor, the remaining amount, concentration and storage conditions of the agent in the storage tank are monitored in real time, the storage tank status data is generated, and the storage tank status data is transmitted to the control module through the wireless transmission module. The control module generates an assist instruction based on the storage tank status data, and the assist instruction is used to generate a coordination instruction in combination with the control instruction, and the spraying module is controlled to spray the agent based on the coordination instruction.

9. The intelligent agent spraying decoking system for energy-saving decoking in the furnace according to claim 8, characterized in that: The storage tank is connected to the spraying module through a corrosion-resistant pipe and is equipped with a medicine pump, and the medicine pump is used to deliver the medicine according to the assistance instruction in the coordination instruction of the control module.

Images