Ash and scale cleaning control method and device for tail flue and electronic equipment

Abstract

The embodiment of the invention discloses an ash cleaning control method and device for a tail flue and electronic equipment. The method comprises the following steps: inputting parameter data and blending combustion fuel components into a trained ash scale prediction model to obtain an ash scale grade output by the ash scale prediction model; in response to the fact that the ash scale grade is higher than the initial grade, spectral data of the tail flue are obtained, the spectral data and the blending combustion fuel components are input into the trained component analysis model, and the ash scale layer number output by the component analysis model and the ash scale component corresponding to each layer are obtained; and determining ash removal operation corresponding to each ash scale component based on the ash scale grade, so as to execute each ash removal operation in sequence according to the sequence of the ash scale components from the outer layer to the inner layer. In the embodiment of the invention, the cooperative control ash removal of various ash removal operations is realized, the ash scales can be sequentially cleaned from outside to inside, various types of ash scales can be cooperatively cleaned, and the ash scale cleaning effect of the tail flue is ensured.

Description

Technical Field

[0001] The embodiments in this specification belong to the field of pipeline ash and scale cleaning technology, and specifically relate to a method, device and electronic equipment for controlling the cleaning of ash and scale in tail flues. Background Technology

[0002] During the co-combustion of solid waste (such as municipal solid waste, sludge, and industrial hazardous waste), especially when co-combusted with coal or other fuels, the flue gas composition becomes more complex. For example, alkali metals (K, Na) and chlorine (Cl) form low-melting-point chlorides; sulfur (S) reacts with alkali metals to form sulfates; and heavy metals (Pb, Zn, etc.) volatilize and condense in the low-temperature zone. These substances condense, adsorb, and chemically react on the low-temperature surfaces (below their dew point or melting point) of the tail flue (including superheaters, economizers, air preheaters, dust collectors, etc.), forming highly adhesive and corrosive ash. Traditional methods of ash removal are relatively simple (such as steam blowing), making it difficult to clean multiple types of ash in the flue simultaneously, resulting in poor cleaning effects and potentially exacerbating caking. Summary of the Invention

[0003] Embodiments of this disclosure provide a method, apparatus, and electronic device for controlling the cleaning of ash and dirt in tail flues, aimed at solving one or more of the above-mentioned problems and other potential problems.

[0004] According to a first aspect of this disclosure, a method for controlling the cleaning of ash and scale in a tail flue is provided. The method includes acquiring parameter data and fuel composition of the tail flue; inputting the parameter data and fuel composition into a trained ash and scale prediction model to obtain the ash and scale level output by the prediction model; the parameter data includes temperature, pressure, temperature change value and pressure change value within a preset time period; in response to the ash and scale level being higher than the initial level, acquiring spectral data of the tail flue; inputting the spectral data and fuel composition into a trained component analysis model to obtain the number of ash and scale layers and the corresponding ash and scale composition of each layer output by the component analysis model; and determining the cleaning operation corresponding to each ash and scale composition based on the ash and scale level, and performing each cleaning operation sequentially according to the ash and scale composition from the outer layer to the inner layer.

[0005] According to a second aspect of this disclosure, a tail flue ash cleaning control device is provided. The device includes a first model processing module configured to acquire parameter data and blended fuel composition of the tail flue, input the parameter data and blended fuel composition into a trained ash prediction model, and obtain the ash level output by the ash prediction model. The parameter data includes temperature, pressure, temperature change value and pressure change value within a preset time period. A second model processing module configured to acquire spectral data of the tail flue in response to the ash level being higher than the initial level, input the spectral data and blended fuel composition into a trained component analysis model, and obtain the number of ash layers and the corresponding ash composition of each layer output by the component analysis model. A ash cleaning control module configured to determine the ash cleaning operation corresponding to each ash composition based on the ash level, and to execute each ash cleaning operation sequentially according to the ash composition order from the outer layer to the inner layer.

[0006] According to a third aspect of this disclosure, an electronic device is provided, including one or more processors and a memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform a method provided according to a first scheme.

[0007] According to a fourth aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the method provided according to the first aspect.

[0008] The solution provided in the embodiments of this specification can determine the ash level, number of ash layers, and composition of each ash layer in the tail flue according to the ash prediction model and composition analysis module. This allows for the determination of the cleaning operation for the ash composition at the corresponding ash level, and the execution of the corresponding cleaning operation layer by layer in sequence according to the number of ash layers. This achieves coordinated control of multiple cleaning operations and can clean the ash in sequence from the outside to the inside, enabling coordinated cleaning of multiple types of ash and ensuring the cleaning effect of the tail flue. Attached Figure Description

[0009] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. In the drawings, the same or similar reference numerals denote the same or similar elements, wherein:

[0010] Figure 1 A schematic flowchart of a method for controlling the cleaning of ash and dirt in the tail flue according to some embodiments of the present disclosure is shown;

[0011] Figure 2 A schematic diagram of the structure of a tail flue cleaning control device according to some embodiments of the present disclosure is shown;

[0012] Figure 3A schematic block diagram of an electronic device according to some embodiments of the present disclosure is shown. Detailed Implementation

[0013] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0014] The terms “comprising” and “having”, and any variations thereof, in this specification, claims, and the foregoing drawings are intended to cover a non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. Depending on the context, the word “if” as it applies herein may be interpreted as “when”, “in response to determination”, or “in response to detection”.

[0015] Figure 1 A schematic flowchart of a tail flue cleaning control method 100 according to some embodiments of the present disclosure is shown. Method 100 can be executed by a terminal, which may include, but is not limited to, a mobile phone, tablet computer, desktop computer, server, etc. Figure 1 As shown in box 102, method 100 can obtain parameter data and blended fuel composition of the tail flue, input the parameter data and blended fuel composition into the trained ash and scale prediction model, and obtain the ash and scale level output by the ash and scale prediction model. The parameter data includes temperature, pressure, temperature change value and pressure change value within a preset time.

[0016] In this embodiment, the parameter data and co-combustion components corresponding to the tail flue are first acquired. The temperature and pressure parameters can be directly collected by temperature and pressure sensors installed in the tail flue. Temperature and pressure changes can be calculated based on the collected temperature and pressure changes over a preset time period (e.g., 10 minutes). The co-combustion components are the components of the solid waste added for co-combustion within a certain time period, which can be determined directly by querying the co-combustion production schedule. Considering that the presence of ash in the tail flue will lead to a temperature increase, pressure rise, and even sudden temperature and pressure changes, the more pronounced these conditions, the more severe the ash situation. Therefore, the ash situation can be judged using parameter data. Furthermore, different co-combustion fuel components will result in different products and different amounts of heat generated during combustion. To avoid misjudgments, the judgment process for different co-combustion fuel components needs to be differentiated. Therefore, an ash prediction model can be pre-trained to predict the ash level based on parameter data and co-combustion fuel components, predicting the ash level under the current condition. Ash and scale prediction models can be trained using machine learning models such as decision trees and support vector machines. Before making predictions based on input data, the model preprocesses the input data, such as handling missing values, outliers, and normalizing the parameter data, and encoding the blended fuel components. Encoding can be done using unique thermal encoding or by pre-setting corresponding codes for each component. Ash and scale levels represent the severity of ash and scale buildup; higher levels indicate greater severity. As an example, ash and scale levels can be divided differently according to actual needs. For instance, historical blending data for each blended fuel component can be obtained to determine the area / thickness of ash and scale actually detected in each historical blending data point. Taking thickness as an example, the thickness range obtained from summarizing all historical blending data can be divided into several thickness intervals, each corresponding to an ash and scale level (generally, the thicker the thickness, the higher the corresponding ash and scale level). During the training of the ash and scale prediction model, the corresponding parameter data and blended fuel components are labeled with the corresponding ash and scale levels based on the thickness intervals as training data. If there is no ash buildup, the ash buildup level can be set to the default initial level (e.g., level 0) to represent the absence of ash buildup. Furthermore, since the tail flue may be quite long, sensors will be installed at multiple locations to collect data. In this case, the tail flue can also be segmented according to the sensor layout, and the ash buildup status of each segment can be assessed separately.

[0017] In box 104, method 100 can, in response to a ash level higher than the initial level, acquire spectral data of the tail flue, input the spectral data and the blended fuel composition into the trained composition analysis model, and obtain the number of ash layers and the corresponding ash composition of each layer output by the composition analysis model.

[0018] In this embodiment, if the ash level is higher than the initial level, it is considered that ash is likely present in the tail flue. At this time, spectral data from the tail flue will be collected using, for example, an optical probe installed in the tail flue. This spectral data will then be analyzed using a trained component analysis model. To improve the accuracy of the model's output, the component analysis model also considers the blended fuel components during training, helping the model to more accurately determine the various ash components. Ultimately, the ash composition needs to be determined by combining both spectral data and blended fuel components. The component analysis model can use machine learning models such as support vector regression or random forest regression, or classical chemometric models such as multiple linear regression, principal component regression, or partial least squares regression. Before making predictions based on the input data, the component analysis model will preprocess the input data, such as performing baseline correction, smoothing and denoising, dimensionality reduction (e.g., recursive feature elimination), and vector normalization on the spectral data, and encoding the blended fuel components. The spectral data corresponding to different ash components have significantly different characteristics. There are also obvious discontinuities between different components in continuous spectral data. These characteristics are different from those of the tail flue wall. Therefore, each ash component can be identified based on the spectral data. Furthermore, the number of ash layers can be determined based on the arrangement order of different ash components (generally, one ash component corresponds to one layer, but the actual ash is formed by the superposition of multiple ash components).

[0019] In box 106, method 100 can determine the cleaning operation corresponding to each dirt component based on the dirt level, and perform each cleaning operation sequentially according to the dirt component order from the outer layer to the inner layer.

[0020] In this embodiment, different dirt components require different cleaning methods. For example, loose fly ash layers can be cleaned using sonic cleaning, while dirt with a certain viscosity can be cleaned using steam blowing. Therefore, corresponding cleaning operations can be pre-set for each dirt component, and each dirt component can correspond to more than one cleaning operation. An applicable dirt level range can be set for each cleaning operation to allow selection under different dirt levels. For example, for relatively hard dirt, when the dirt level is low (i.e., the dirt is not severe), shock wave blowing can be used, processed by a shock wave generator. However, when the dirt level is high (i.e., the dirt is severe), the dirt may be extremely hard and difficult to clean with shock waves. In this case, mechanical assisted cleaning can be used, with a retractable micro-scraper scraping and cleaning. The threshold for distinguishing whether the dirt level is high or low can be adjusted. For example, the highest level can be divided by two and rounded down, with the rounded value used as the level threshold. Ultimately, the cleaning operation required for each identified dirt component will be determined based on the dirt level. The cleaning operation will be executed sequentially from the outermost layer to the innermost layer, corresponding to the dirt component in each layer. This collaborative cleaning operation, starting from the outermost layer, progressively selects the most suitable operation to clean the dirt layer by layer. Compared to a single-dimensional cleaning process, this method is more adapted to the actual composition and layout of the dirt, resulting in better cleaning effects and reducing the need for repeated cleaning due to poor cleaning results, or situations where dirt cannot be cleaned at all, or even worsens dirt buildup. Furthermore, this collaborative cleaning control method can adaptively adjust the cleaning strategy, intensity, and frequency according to the specific characteristics of the identified dirt, enabling better coordinated cleaning of multiple types of dirt.

[0021] In one possible implementation, the cleaning operation corresponding to each type of dirt component is determined based on the dirt level, including:

[0022] Query at least one optional operation for each dirt component, and determine the target optional operation that matches the recommended level and dirt level among the optional operations, so as to determine the target optional operation as the cleaning operation for the dirt component.

[0023] In this embodiment, based on historical co-firing data, the types of various ash components that may be generated during the co-firing process (e.g., loose, hard, bonded, corrosive, etc.) can be determined, and corresponding optional operations are set for each ash component (e.g., optional operations for loose ash components may include sonic cleaning and compressed air blowing; optional operations for hard ash components may include shock wave blowing and mechanically assisted cleaning; optional operations for bonded ash components may include steam blowing and chemical cleaning; optional operations for corrosive ash components may include chemical cleaning and mechanically assisted cleaning, etc.). Each optional operation has a pre-set recommended level based on experience; that is, this operation is only recommended after the ash level reaches the recommended level, in order to reduce cleaning energy consumption and avoid damaging the inner wall by using overly harsh cleaning methods at low levels. Where only one optional operation is set for a ash component, that optional operation can default to all levels, meaning that this operation is used for cleaning at all levels. By querying the database to identify the corresponding optional operations for the identified dirt components, and then determining the most suitable target optional operation from among the optional operations based on the dirt level, these target optional operations are the cleaning operations used for this cleaning of the dirt components. The method for matching the recommended level with the dirt level can be to calculate the Pearson correlation coefficient, and the optional operation with the highest correlation coefficient is selected as the target optional operation.

[0024] In one possible implementation, acquiring spectral data of the tail flue includes:

[0025] The spectral probe is controlled to scan the tail flue along a preset path to obtain spectral data of the tail flue, and the spatial coordinates of the spectral probe when acquiring each spectral data are recorded.

[0026] By integrating the spectral data based on spatial coordinates, a spectral distribution map of the tail flue is obtained;

[0027] The method also includes:

[0028] Based on the spectral coordinates of the identified ash components in the spectral distribution map, the coordinates of the ash in the tail flue are determined.

[0029] In this embodiment, to acquire continuous spectral data and facilitate subsequent determination of the location of identified ash and dirt within the channel, a fixed rotation / movement path can be pre-set for the spectral probe. The probe scans the tail flue according to this preset path, and during scanning, the spatial coordinates and / or the azimuth angle pointed to by the probe are marked. This allows for the integration and stitching of spectral data at various spatial coordinates / azimuth angles in sequence along the path, generating a spectral distribution map of the tail flue. This map represents the overall spectral data distribution within a certain scanned range of the tail flue. When the component analysis model identifies ash and dirt components, the specific location of the ash and dirt within the tail flue can be determined based on the spectral coordinates corresponding to the spectral data of the identified ash and dirt component in the spectral distribution map and the ash and dirt coordinates mapped from these coordinates. Thus, during cleaning operations, if the corresponding pipe / nozzle allows for angle control, the angle can be controlled to orient the pipe / nozzle towards the ash and dirt coordinates, improving the cleaning effect.

[0030] In one possible implementation, the cleaning operation includes shock wave cleaning, acoustic cleaning, steam cleaning, compressed air purging, chemical cleaning, and mechanically assisted cleaning.

[0031] In this embodiment, shockwave soot blowing is a method of breaking and peeling off ash by igniting a mixed gas to generate a controlled explosion, forming a shock wave of a certain intensity, which acts on the inertial and shear forces of the ash particles, causing them to break apart and peel off. Acoustic soot blowing uses a high-intensity acoustic generator to produce low-frequency, high-energy sound waves, causing air molecules and ash particles to oscillate at high frequencies, disrupting the binding force and electrostatic force between the ash particles, putting them in a "suspended" state, and allowing them to be carried away by the flue gas. Steam soot blowing / compressed air purging uses a high-speed jet of high-pressure steam / compressed air to purge the heated surface at specific points and times. Chemical soot blowing involves injecting trace amounts of chemical reagents (such as neutralizing agents or loosening agents) to weaken the adhesion of the inner layer of ash through chemical reactions, assisting in physical soot removal. Mechanically assisted soot blowing integrates micro-mechanical scraping or vibration devices (such as retractable scrapers) to handle hardened ash particles under low load. The cleaning equipment / pipes corresponding to the above-mentioned cleaning operations can all have fixed inlet points set in the tail flue to connect to the tail flue through specific pipes. Generally, they can be arranged in sections at a distance from the vertical flue.

[0032] In one possible implementation, the method further includes:

[0033] Based on historical blending data, a training set was determined. The training set included data pairs constructed from parameter data samples and blended fuel composition samples, as well as ash and scale level samples.

[0034] Based on the data pairs, the predicted dirt level is generated from the initial model;

[0035] Using the dirt and grime level samples as supervision signals, the initial model is trained for at least one round to obtain the dirt and grime prediction model.

[0036] In this embodiment, based on historical blending data, samples of blended fuel components used in historical blending tasks and parameter data samples collected during the blending process can be read. Furthermore, the historical blending data may also include the ash and scale detected in the tail flue after blending of the corresponding fuel components. Different ash and scale grade samples can be manually assigned based on the area, number of layers, and thickness of the ash and scale. This allows the matched parameter data samples and blended fuel component samples to be associated, forming data pairs. A training set for the model is then constructed based on these data pairs and the ash and scale grade samples. During the initial model training using the training set, the model's generator can generate predicted ash and scale grades based on the data pairs. The generator loss is obtained by comparing the ash and scale grade samples with the predicted ash and scale grades. This generator loss is then used to assess the loss of the predicted ash and scale grades. The loss assessment can utilize a comparison loss function (e.g., labeled smoothed cross-entropy loss function). The generator loss can be a relatively large value, and then it can be backpropagated to the generator to guide the optimization of the generator's parameters, achieving a round of supervised training of the generator. This training process can be executed iteratively round after round until the generator can generate more accurate predictions of dirt levels, that is, until the loss value calculated by the loss function is smaller. After training, the dirt prediction model can output the dirt level.

[0037] Furthermore, the training process of the component analysis model can be similar to that of the ash and dirt prediction model. The spectral data samples and the blended fuel component samples are associated as the first data pair, and the ash and dirt layer number samples and the ash and dirt component samples are associated as the second data pair. The first data pair is used as the input data of the model, and the second data pair is used as the output data of the model to train the component analysis model.

[0038] In one possible implementation, the method further includes:

[0039] The estimated thickness of each dirt component is determined based on spectral data, the estimated cleaning time for each dirt component is determined based on the estimated thickness, and the maximum execution time of the cleaning operation is set based on the estimated cleaning time.

[0040] In this embodiment, when training the component analysis model, the training set can also include estimated thickness samples of dirt components in the output data samples, enabling the trained model to predict the estimated thickness of each dirt component. The estimated thickness samples can be determined by combining an ultrasonic thickness gauge with the number of signal points between boundary layers where the spectral signal changes drastically (i.e., indicating a change from one component to another) in the spectral data. Specifically, the average of the first thickness measured by the ultrasonic thickness gauge and the second thickness determined based on the number of signal points in the spectral data can be used as the estimated thickness sample. Thus, the trained component analysis model can also determine the estimated thickness of each dirt component based on the spectral data. During the aforementioned process of determining the cleaning operation, each cleaning operation can be pre-set with an estimated cleaning efficiency (i.e., how much dirt volume can be cleaned per second) based on historical cleaning data and experience. Combining the estimated thickness and estimated area to calculate the estimated volume, the estimated cleaning time for that dirt component can be determined. This allows for setting the maximum execution time of the corresponding cleaning operation, ensuring that the cleaning operation's execution time does not exceed the maximum execution time and reducing ineffective cleaning. The estimated area can be determined, for example, by acquiring images of the dirt using a camera probe and then performing image recognition on the images.

[0041] In one possible implementation, after setting the maximum execution time of the dust removal operation based on the estimated cleaning time, the method further includes:

[0042] When the execution time of the currently executing cleaning operation reaches the maximum execution time, the next cleaning operation is executed, and the spectral data is reacquired;

[0043] Based on the reacquired spectral data, the number of dirt layers and the estimated thickness of the outermost dirt layer were re-determined.

[0044] In response to the fact that the redefined number of dirt layers is unchanged from the initially obtained number of dirt layers, and the estimated thickness of the outermost dirt component is greater than the thickness threshold, the target cleaning operation corresponding to the next dirt level for the outermost dirt component is determined and the target cleaning operation is executed.

[0045] In this embodiment, the maximum execution time can generally be set with a certain additional margin. Therefore, after the execution time of the currently executing cleaning operation reaches the maximum execution time, it is generally considered that the corresponding dirt layer should have been cleaned. At this time, the next cleaning operation will be executed sequentially, and spectral data can be reacquired. If the number of dirt layers analyzed from the reacquired spectral data is unchanged from before (i.e., the number of dirt layers has not decreased), and the estimated thickness of the outermost dirt component is greater than the thickness threshold, then the cleaning effect of the just completed cleaning operation is considered poor, and this layer of dirt has not been cleaned well, with a large amount of dirt residue remaining. At this time, the next cleaning operation that was in progress can be directly interrupted. Among the cleaning operations corresponding to the dirt components, the target cleaning operation corresponding to the next dirt level of the current dirt level is determined. In this way, the outermost dirt component is cleaned again with a cleaning method with higher cleaning intensity. This achieves real-time determination of the dirt status during the cleaning process and timely adjustment of the cleaning operation to ensure the cleaning effect on the dirt.

[0046] Figure 2 A schematic diagram of the structure of a tail flue ash and dirt cleaning control device 200 according to some embodiments of this disclosure is shown. The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for the device embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and relevant parts can be referred to the description of the method embodiments. Figure 2 As shown, the device 200 includes a first model processing module 201, configured to acquire parameter data and blended fuel composition of the tail flue, input the parameter data and blended fuel composition into a trained ash prediction model, and obtain the ash level output by the ash prediction model. The parameter data includes temperature, pressure, temperature change value and pressure change value within a preset time period; a second model processing module 202, configured to acquire spectral data of the tail flue in response to the ash level being higher than the initial level, input the spectral data and blended fuel composition into a trained component analysis model, and obtain the number of ash layers and the corresponding ash composition of each layer output by the component analysis model; and a ash removal control module 203, configured to determine the ash removal operation corresponding to each ash composition based on the ash level, and execute each ash removal operation sequentially according to the ash composition order from the outer layer to the inner layer.

[0047] In one possible implementation, the dust removal control module 203 is further configured to query at least one optional operation corresponding to each dust component, and determine a target optional operation whose recommended level matches the dust level among the optional operations, so as to determine the target optional operation as the dust removal operation for the dust component.

[0048] In one possible implementation, the second model processing module 202 is further configured to control the spectral probe to scan the tail flue along a preset path to obtain the spectral data of the tail flue and record the spatial coordinates when the spectral probe collects each spectral data; integrate the spectral data according to the spatial coordinates to obtain the spectral distribution map of the tail flue; and determine the ash coordinates of the tail flue based on the spectral coordinates corresponding to the identified ash components in the spectral distribution map.

[0049] In one possible implementation, the cleaning operation includes shock wave cleaning, acoustic cleaning, steam cleaning, compressed air purging, chemical cleaning, and mechanically assisted cleaning.

[0050] In one possible implementation, the apparatus further includes a model training module configured to determine a training set based on historical blending data, the training set including data pairs constructed from parameter data samples and blended fuel composition samples, and ash and dirt level samples; generate predicted ash and dirt levels from an initial model based on the data pairs; and perform at least one round of model training on the initial model using the ash and dirt level samples as supervision signals to obtain an ash and dirt prediction model.

[0051] In one possible implementation, the device further includes a thickness calculation module configured to determine the estimated thickness of each dirt component based on spectral data, to determine the estimated cleaning time of the dirt component based on the estimated thickness, and to set the maximum execution time of the cleaning operation based on the estimated cleaning time.

[0052] In one possible implementation, the thickness calculation module is further configured to, in response to the current cleaning operation reaching its maximum execution time, execute the next cleaning operation and reacquire spectral data; based on the reacquired spectral data, redetermine the number of dirt layers and the estimated thickness of the outermost dirt component; and in response to the redetermined number of dirt layers being unchanged relative to the initially acquired number of dirt layers, and the estimated thickness of the outermost dirt component being greater than a thickness threshold, determine the target cleaning operation corresponding to the next dirt level for the outermost dirt component and execute the target cleaning operation.

[0053] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this specification is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in or transmitted through a computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., Digital Versatile Discs (DVDs)), or semiconductor media (e.g., Solid State Disks (SSDs)).

[0054] Figure 3 A block diagram of an electronic device 300 that can implement various embodiments of the present disclosure is shown. For example... Figure 3 As shown, the electronic device 300 includes a processor 310, a disk drive 320, an input / output interface 330, a network interface 340, and a memory 350. The processor 310, disk drive 320, input / output interface 330, network interface 340, and memory 350 can communicate with each other via a communication bus 360.

[0055] The processor 310 can be implemented using a general-purpose CPU, microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits to execute relevant programs in order to implement the technical solution provided in this application.

[0056] The memory 350 can be implemented in the form of ROM (Read Only Memory), RAM (Read Access Memory), static memory, dynamic storage devices, etc. The memory 350 can store the operating system 351 used to control the operation of the electronic device 300, and the basic input / output system (BIOS) 352 used to control the low-level operations of the electronic device 300. Additionally, it can store a web browser 353, a data storage management system 354, etc. In summary, when the technical solution provided in this application is implemented through software or firmware, the relevant program code is stored in the memory 350 and is called and executed by the processor 310.

[0057] Input / output interface 330 is used to connect input / output modules to realize information input and output. Input / output modules can be configured as components in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touch screens, microphones, various sensors, etc., and output devices may include displays, speakers, vibrators, indicator lights, etc.

[0058] Network interface 340 is used to connect a communication module (not shown in the figure) to enable communication and interaction between the device and other devices. The communication module can communicate via wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0059] Bus 360 includes a pathway for transmitting information between various components of the device, such as processor 310, disk drive 320, input / output interface 330, network interface 340, and memory 350.

[0060] It should be noted that although the above-described device only shows the processor 310, disk drive 320, input / output interface 330, network interface 340, memory 350, bus 360, etc., in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the method of this application, and does not necessarily include all the components shown in the figures.

[0061] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0062] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing. Furthermore, although operations are depicted in a specific order, this should be understood as requiring that such operations be performed in the specific order shown or in sequential order, or requiring that all illustrated operations be performed to achieve the desired result. In certain environments, multitasking and parallel processing may be advantageous. Similarly, while several specific implementation details are included in the foregoing discussion, these should not be construed as limiting the scope of this disclosure. Certain features described in the context of individual embodiments may also be implemented in combination in a single implementation. Conversely, various features described in the context of a single implementation may also be implemented individually or in any suitable sub-combination in multiple implementations.

[0063] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative examples of implementing the claims.

Claims

1. A method of ash cleaning control of a tail flue, characterized by, The method comprises: obtaining parameter data and fuel blending component of the tail flue, inputting the parameter data and fuel blending component into the trained ash prediction model to obtain the ash level output by the ash prediction model, and the parameter data comprises temperature, pressure, temperature change value and pressure change value within a preset time length; in response to the ash level being higher than the initial level, obtaining spectral data of the tail flue, inputting the spectral data and fuel blending component into the trained component analysis model to obtain the ash layer number and corresponding ash component of each layer output by the component analysis model; determining the ash removal operation corresponding to each ash component based on the ash level, and sequentially executing each ash removal operation in the order of ash component from the outer layer to the inner layer.

2. A method of ash cleaning control of a tail flue according to claim 1, characterized in that, The determination of the ash removal operation corresponding to each ash component based on the ash level comprises: querying at least one optional operation corresponding to each ash component, determining a target optional operation in which the recommended level matches the ash level, and determining the target optional operation as the ash removal operation of the ash component.

3. A method of ash cleaning control of a tail flue according to claim 1, characterized in that, The obtaining of the spectral data of the tail flue comprises: controlling the spectral probe to scan the tail flue according to a preset path to obtain the spectral data of the tail flue, and recording the spatial coordinates of the spectral probe when collecting each spectral data; integrating each spectral data according to the spatial coordinates to obtain a spectral distribution map of the tail flue; The method further comprises: determining the ash coordinates of the tail flue based on the spectral coordinates corresponding to the identified ash component in the spectral distribution map.

4. A method of ash cleaning control of a tail flue according to claim 1, wherein The ash removal operation comprises shock wave soot blowing operation, acoustic soot blowing operation, steam soot blowing operation, compressed air blowing operation, chemical soot blowing operation and mechanical auxiliary soot blowing operation.

5. A method of ash cleaning control of a tail flue according to claim 1, wherein The method further comprises: determining a training set based on historical blending data, the training set comprising data pairs constructed by parameter data samples and fuel blending component samples, and ash level samples; generating a predicted ash level from an initial model based on the data pairs; using the ash level samples as a supervision signal, performing at least one round of model training on the initial model to obtain an ash prediction model.

6. A method of ash cleaning control of a tail flue according to claim 1, wherein The method further comprises: determining the estimated thickness of each ash component based on the spectral data, determining the estimated cleaning time length of the ash component according to the estimated thickness, and setting the maximum execution time length of the ash removal operation according to the estimated cleaning time length.

7. A method of ash cleaning control of a tail flue according to claim 6, characterized in that, After setting the maximum execution time length of the ash removal operation according to the estimated cleaning time length, the method further comprises: in response to the execution time length of the currently executed ash removal operation reaching the maximum execution time length, executing the next ash removal operation and reacquiring the spectral data; redetermining the ash layer number and the estimated thickness of the outermost ash component based on the reacquired spectral data; in response to the redetermined ash layer number being unchanged relative to the initially acquired ash layer number and the estimated thickness of the outermost ash component being greater than a thickness threshold, determining the target ash removal operation corresponding to the next ash level of the outermost ash component, and executing the target ash removal operation.

8. A device for controlling the cleaning of ash deposits from a tail flue, characterized in that The device comprises: The first model processing module is configured to obtain parameter data of the tail flue and a blending fuel component, input the parameter data and the blending fuel component into a trained ash prediction model, and obtain an ash level output by the ash prediction model, wherein the parameter data includes temperature, pressure, a temperature change value and a pressure change value within a preset time period; The second model processing module is configured to, in response to the ash level being higher than an initial level, obtain spectrum data of the tail flue, input the spectrum data and the blending fuel component into a trained component analysis model, and obtain an ash layer number and a corresponding ash component of each layer output by the component analysis model; The ash cleaning control module is configured to determine, based on the ash level, an ash cleaning operation corresponding to each ash component, and sequentially execute each ash cleaning operation in order of ash component ranking from an outer layer to an inner layer.

9. An electronic device, comprising: Comprising: one or more processors, a memory associated with the one or more processors, the memory for storing program instructions that, when read and executed by the one or more processors, perform the steps of the ash cleaning control method of the tail flue of any one of claims 1-7.

10. Computer program product, characterized in that, A computer program that, when executed by a processor, implements the ash cleaning control method of the tail flue according to any one of claims 1-7.

Images