Coal-fired boiler low-load stable combustion monitoring control device and method

Abstract

The invention discloses a coal-fired boiler low-load stable combustion monitoring control device and method, and solves the problems that when a combustion stable combustion index of a coal-fired boiler is determined according to a flame combustion state of the coal-fired boiler in an existing method, the combustion stable combustion index is influenced by real-time load parameters and boiler operation state data, and the combustion stable combustion index is influenced. The device comprises a data acquisition module, a load prediction module, an optimization control module and a stable combustion execution module, and the stable combustion execution module comprises an equipment base, a preprocessing base fixedly installed on the equipment base and a fuel release pipe. The system is provided with the load prediction module, and the load prediction module constructs a load prediction model by combining a conditional value-at-risk method and a long-short-term memory neural network model, thereby effectively combining real-time load parameters and boiler operation state data, and achieving the precise prediction of the load power. And therefore, the low-load stable-combustion optimized operation of the coal-fired boiler is ensured.

Description

Technical Field

[0001] The invention belongs to the technical field of coal-fired boiler power generation, and in particular relates to a low-load stable combustion monitoring and control device and method for a coal-fired boiler. Background Art

[0002] With the continuous optimization of the power energy structure, the proportion of renewable energy such as hydropower and photovoltaics in the power generation field has gradually increased. This trend provides strong support for the green transformation of the energy structure. However, the inherent characteristics of renewable energy, including its intermittent, volatile and random nature affected by weather, seasons and day and night changes, make its absorption and peak-shaving capabilities have obvious shortcomings compared with traditional thermal power generation. In order to effectively solve the problem of new energy absorption and ensure the stability of power supply, thermal power units must have more excellent deep peak-shaving capabilities.

[0003] Chinese patent CN118149356A discloses a method and system for controlling low-load stable combustion of a coal-fired boiler, the method comprising: determining a stable combustion index of the coal-fired boiler according to the flame combustion state of the coal-fired boiler, the stable combustion index being used to indicate the combustion stability of the coal-fired boiler; when the operating load of the coal-fired boiler is less than a first preset threshold value, a boiler air-powder quantity feedforward value can be given according to the result of combustion optimization, and the furnace combustion stability can be adjusted under steady-state conditions, or the combustion stability judgment index can be used to perform feedback control of the furnace combustion under abnormal conditions and boiler variable conditions; however, when the existing method determines the stable combustion index of the coal-fired boiler by the flame combustion state of the coal-fired boiler, the stable combustion index is affected by the real-time load parameters and the boiler operating state data, and a single flame combustion state cannot fully reflect the combustion process. In view of the above problems, we propose a low-load stable combustion monitoring and control device and method for a coal-fired boiler. Summary of the invention

[0004] The purpose of the present invention is to address the deficiencies in the prior art and provide a low-load stable combustion monitoring and control device and method for a coal-fired boiler, thereby solving the problem that when the prior method determines the combustion and stable combustion index of a coal-fired boiler through the flame combustion state of the coal-fired boiler, the combustion and stable combustion index is affected by the real-time load parameters and the boiler operation status data, and a single flame combustion state cannot fully reflect the combustion process.

[0005] When the existing method determines the combustion and stabilization index of a coal-fired boiler through the flame combustion state of the coal-fired boiler, the combustion and stabilization index is affected by the real-time load parameters and the boiler operation status data. A single flame combustion state cannot fully reflect the combustion process. In response to the above problems, we propose a low-load stable combustion monitoring and control device and method for a coal-fired boiler. In short, the control device consists of a data acquisition module, a load prediction module, an optimization control module, and a stable combustion execution module. When working, the data acquisition module first obtains the operation status data and real-time load value of the coal-fired boiler, and then the load prediction module, combining the conditional risk value method and the long short-term memory neural network model, constructs a load prediction model. The load prediction model identifies the operation status data and the real-time load value, and outputs the load power prediction sequence. Then the optimization control module calculates the stable combustion parameters of the coal-fired boiler based on the load power prediction sequence, and superimposes the coal-fired boiler stable combustion parameters with the load power prediction sequence to obtain a predicted stable combustion optimization instruction. Finally, the stable combustion execution module responds to the predicted stable combustion optimization instruction and performs the stable combustion optimization work. In an embodiment of the present invention, a load prediction module is provided. The load prediction module constructs a load prediction model by combining the conditional risk value method and the long short-term memory neural network model, thereby effectively combining the real-time load parameters and the boiler operation status data, and realizing the accurate prediction of the load power, thereby ensuring the low-load stable combustion optimization operation of the coal-fired boiler, and overcoming the problem that when the existing method determines the combustion stability index of the coal-fired boiler through the flame combustion state of the coal-fired boiler, the combustion stability index is affected by the real-time load parameters and the boiler operation status data, and a single flame combustion state cannot fully reflect the combustion process.

[0006] The present invention is implemented in this way: a low-load stable combustion monitoring and control device for a coal-fired boiler, the low-load stable combustion monitoring and control device for a coal-fired boiler comprises:

[0007] Data acquisition module, used to obtain the operating status data and real-time load value of the coal-fired boiler;

[0008] The load forecasting module combines the conditional value at risk method and the long short-term memory neural network model to build a load forecasting model. The load forecasting model identifies the operating status data and the real-time load value and outputs the load power forecast sequence.

[0009] An optimization control module is used to obtain a load power prediction sequence, calculate the stable combustion parameters of the coal-fired boiler based on the load power prediction sequence, and superimpose the stable combustion parameters of the coal-fired boiler with the load power prediction sequence to obtain a prediction stable combustion optimization instruction;

[0010] The combustion stabilization execution module executes the combustion stabilization optimization work in response to the predicted combustion stabilization optimization instruction.

[0011] Optionally, the load forecasting module includes:

[0012] A data processing unit is used to obtain operation status data and real-time load values, and decompose the operation status data to obtain a state decomposition set;

[0013] The method for decomposing and processing the running status data specifically includes:

[0014] Based on the empirical mode decomposition algorithm, the operating status data is decomposed to obtain a random oscillation sequence;

[0015]

[0016] q represents the operating status data decomposition coefficient, u i is the random oscillation sequence during decomposition, Q n represents the decomposition trend term, and x(t) represents the input operation status data;

[0017]

[0018] maxx(t) and minx(t) represent the local maximum and local minimum values ​​in the input running status data, respectively. Represents the local mean of the input operating status data;

[0019] Obtain a random oscillation sequence, introduce white noise with an amplitude of δ and an expansion factor α into the random oscillation sequence, reconstruct the random oscillation sequence, and obtain a reconstructed oscillation sequence;

[0020] The reconstructed oscillation sequence output is expressed as:

[0021]

[0022] Calculate the overall mean of the reconstructed oscillation sequence and use the overall mean as the state decomposition set;

[0023] A model definition unit is used to load a state decomposition set and a real-time load value, and to train a load forecasting model based on a conditional value at risk method and a long short-term memory neural network model based on the state decomposition set and the real-time load value, and output the load forecasting model;

[0024] The prediction execution unit, the load prediction model identifies the operating status data and the real-time load value, and outputs the load power prediction sequence.

[0025] Optionally, the load forecasting model based on the state decomposition set and the real-time load value training based on the conditional value at risk method and the long short-term memory neural network model specifically includes:

[0026] The long short-term memory neural network model is used as the initial model. The initial model includes an input layer, a hidden layer, and an output layer. The input layer introduces peak load balance constraints, load power balance constraints, stable combustion burner output constraints, and coal-fired boiler ramp rate constraints.

[0027] The conditional risk value algorithm is introduced into the hidden layer to construct the low load objective function of the load forecasting model;

[0028] Among them, the low load objective function is defined as follows:

[0029]

[0030] y loa (t+1) is the load power forecast sequence for the next cycle, y loa (t) is the current load power, y loa (t-1) is the load power of the previous cycle, w y represents the peak load balance constraint, v t represents the ramp rate constraint of the coal-fired boiler, w t represents the output constraint of the stable combustion burner, represents the rated output constraint of the stable combustion burner, Q r is the total work output of the stable combustion burner, g y is the inertia coefficient of the stable combustion burner;

[0031]

[0032] Obtain a state decomposition set and a real-time load value, and extract characteristic factors corresponding to the state decomposition set and the real-time load value at the input layer to obtain a characteristic factor set;

[0033] The hidden layer uses the low-load objective function as the excitation function, assigns weights to the characteristic factor set, calculates the node output of the hidden layer, and obtains the load power prediction sequence;

[0034] The mean absolute error and root mean square error are used to verify the difference between the load power forecast sequence and the actual sequence, and to verify the accuracy of the model. If the model meets the preset accuracy of the mean absolute error and root mean square error, the load power forecast sequence is output.

[0035] Optionally, the method for calculating the stable combustion parameters of a coal-fired boiler based on a load power prediction sequence specifically includes:

[0036] The load power prediction sequence is loaded, and the stable combustion parameters of the coal-fired boiler are calculated by taking the difference between the load power prediction sequence and the output power of the coal-fired boiler in the stable combustion period as a constraint to tend to the set difference;

[0037]

[0038] Among them, minΔy loa To set the difference constraint, is the load power forecast sequence y within the set period T loa (t+1), β is the stable combustion control command of the coal-fired boiler assigned by the stable combustion burner;

[0039] Obtain stable combustion parameters of the coal-fired boiler, correct the stable combustion parameters of the coal-fired boiler based on a rolling optimization algorithm, and output the corrected stable combustion parameters of the coal-fired boiler.

[0040] Optionally, the combustion stabilization execution module includes:

[0041] Equipment base;

[0042] A pretreatment seat fixedly mounted on the base of the equipment, wherein a pretreatment chamber is arranged in the pretreatment seat, a discharge guide seat is provided at the bottom of the pretreatment seat, and a pretreatment mechanism is arranged in the pretreatment chamber, and the pretreatment mechanism is used to pretreatment the stable combustion fuel;

[0043] A fuel release pipe is arranged on the equipment base, one end of the fuel release pipe is fixedly connected to the discharge guide seat, the other end of the fuel release pipe is provided with a fuel injection port, and an anti-deposition component is arranged in the fuel release pipe.

[0044] Optionally, the pre-processing mechanism comprises:

[0045] A pre-processing motor, wherein the pre-processing motor is fixedly mounted on the equipment base;

[0046] A pre-processing drive unit fixedly connected to an output end of the pre-processing motor;

[0047] A pre-processing linkage part, which is arranged on the pre-processing seat and connected to the pre-processing driving part;

[0048] The fuel refinement part is arranged in the pretreatment chamber, and the fuel refinement part includes a main refinement roller, an auxiliary refinement roller and a turbulent refinement roller. The main refinement roller and the auxiliary refinement roller are respectively connected to the pretreatment drive part, the turbulent refinement roller is connected to the pretreatment linkage part, and the side wall of the turbulent refinement roller is fixedly installed with turbulent crushing teeth.

[0049] Optionally, the pre-processing driving unit includes:

[0050] A driving gear, which is arranged on one side of the pre-processing motor, is fixedly connected to the output end of the pre-processing motor, and a side of the driving gear away from the pre-processing motor is fixedly connected to the end of the main refining roller;

[0051] A driven gear, the driven gear is rotatably mounted on the side wall of the pre-treatment seat, the driven gear is meshed with the driving gear for transmission, and one side of the driven gear is fixedly connected to the end of the auxiliary thinning roller;

[0052] A linkage gear ring, the linkage gear ring is fixedly mounted on the outer side wall of the driven gear, and a driven gear ring meshing with the linkage gear ring for transmission, the driven gear ring is fixedly connected to the pre-treatment linkage part.

[0053] Optionally, the preprocessing linkage unit includes:

[0054] A pretreatment linkage rod, one end of which is fixedly connected to the driven gear ring, and the pretreatment linkage rod is rotatably mounted on the side wall of the pretreatment seat;

[0055] A first rotating wheel, wherein the first rotating wheel is fixedly mounted on the top of the pre-processing linkage rod;

[0056] A second rotating wheel, the second rotating wheel is arranged on one side of the first rotating wheel, and the second rotating wheel is rotatably connected to the first rotating wheel through a transmission belt;

[0057] A pretreatment driven rod, one end of which is fixedly connected to the second rotating wheel, and the other end of which passes through the top wall of the pretreatment seat and is fixedly connected to the turbulence refinement roller.

[0058] Optionally, the anti-deposit component comprises:

[0059] A driving motor, wherein the driving motor is fixedly mounted on a device base;

[0060] A material conveying auger, wherein the material conveying auger is arranged in the fuel release pipe, and one end of the material conveying auger is fixedly connected to the output end of the driving motor;

[0061] an anti-adsorption sleeve rotatably mounted in the fuel release pipe, wherein the anti-adsorption sleeve is rotatably mounted on an end wall of the fuel release pipe;

[0062] A fuel scraping part, which is arranged in the fuel release pipe and is fixedly connected to the end of the feed auger;

[0063] Wherein, the fuel scraping part comprises:

[0064] A scraper linkage rod, the scraper linkage rod is fixedly connected to the end of the feeding auger;

[0065] At least one set of scraper mounting seats, wherein the scraper mounting seats are fixedly sleeved on the scraper linkage rod;

[0066] A scraping rotating arm, the scraping rotating arm is fixedly mounted on the scraping mounting seat, and a sleeve driving roller is fixedly mounted on the scraping rotating arm, the sleeve driving roller is used to drive the anti-adsorption sleeve;

[0067] An adsorbent scraping roller is rotatably mounted on the sleeve driving roller.

[0068] On the other hand, the present invention also provides a coal-fired boiler low-load stable combustion monitoring and control method, which is implemented by the coal-fired boiler low-load stable combustion monitoring and control device described above. The coal-fired boiler low-load stable combustion monitoring and control method specifically includes:

[0069] Obtain the operating status data and real-time load value of coal-fired boilers;

[0070] The load forecasting model is constructed by combining the conditional risk value method and the long short-term memory neural network model. The load forecasting model identifies the operating status data and the real-time load value and outputs the load power forecast sequence.

[0071] Obtain a load power forecast sequence, calculate the stable combustion parameters of the coal-fired boiler based on the load power forecast sequence, and superimpose the stable combustion parameters of the coal-fired boiler with the load power forecast sequence to obtain a forecast stable combustion optimization instruction;

[0072] In response to the predicted combustion stabilization optimization instruction, a combustion stabilization optimization task is performed.

[0073] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0074] In an embodiment of the present invention, a load prediction module is provided. The load prediction module constructs a load prediction model by combining the conditional risk value method and the long short-term memory neural network model, thereby effectively combining the real-time load parameters and the boiler operation status data, and realizing the accurate prediction of the load power, thereby ensuring the low-load stable combustion optimization operation of the coal-fired boiler, and overcoming the problem that when the existing method determines the combustion stability index of the coal-fired boiler through the flame combustion state of the coal-fired boiler, the combustion stability index is affected by the real-time load parameters and the boiler operation status data, and a single flame combustion state cannot fully reflect the combustion process.

[0075] In an embodiment of the present invention, the operating status data is decomposed based on the empirical mode decomposition algorithm, and a random oscillation sequence is reconstructed. By decomposing the operating status data, the performance of the boiler under different operating conditions can be predicted, and the operating parameters can be adjusted accordingly to achieve optimal combustion efficiency.

[0076] In an embodiment of the present invention, a load forecasting model is provided, which uses a long short-term memory neural network model as an initial model and introduces a conditional value at risk algorithm to complete the construction of the load forecasting model, thereby ensuring the prediction accuracy and robustness of the load forecasting model that combines the long short-term memory neural network model and the conditional value at risk algorithm.

[0077] In the embodiment of the present invention, a pretreatment mechanism is provided, which is composed of a pretreatment drive unit, a pretreatment linkage unit, and a fuel refinement unit. The pretreatment drive unit, the pretreatment linkage unit, and the fuel refinement unit cooperate with each other to achieve sufficient refinement of the fuel pulverized coal in the pretreatment chamber. On the one hand, it ensures uniform injection of the fuel from the fuel injection port, and on the other hand, it is also beneficial to the stable combustion and low-load operation of the coal-fired boiler. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 It is a structural schematic diagram of the low-load stable combustion monitoring and control device for a coal-fired boiler provided by the present invention;

[0079] Figure 2 It is a schematic diagram of the implementation flow of the method for decomposing and processing the operating status data provided by the present invention;

[0080] Figure 3 It is a schematic diagram of the implementation process of the load forecasting model based on the state decomposition set, real-time load value training based on the conditional risk value method and the long short-term memory neural network model provided by the present invention;

[0081] Figure 4 It is a schematic diagram of the implementation flow of the method for calculating the stable combustion parameters of a coal-fired boiler based on the load power prediction sequence provided by the present invention;

[0082] Figure 5 It is a structural schematic diagram of the stable combustion execution module provided by the present invention;

[0083] Figure 6 It is a front view of the combustion stabilization execution module provided by the present invention;

[0084] Figure 7 is a top view of the combustion stabilization execution module provided by the present invention;

[0085] Figure 8 It is a structural schematic diagram of the pretreatment mechanism provided by the present invention;

[0086] Fig. 9 is an axonometric diagram of the pre-processing mechanism provided by the present invention;

[0087] Fig.10 It is a front view of the pretreatment mechanism provided by the present invention;

[0088] Fig.11 yes Fig.10 AA section view;

[0089] Fig.12 It is a structural schematic diagram of the fuel refinement unit provided by the present invention;

[0090] Fig.13 It is a structural schematic diagram of the anti-deposit assembly provided by the present invention;

[0091] Fig.14 It is a schematic diagram of the three-dimensional structure of the anti-deposit assembly provided by the present invention;

[0092] Fig.15 is a front view of the anti-deposit assembly provided by the present invention;

[0093] Fig.16 yes Fig.15 BB section view;

[0094] Fig.17 It is a structural schematic diagram of the fuel scraping portion provided by the present invention;

[0095] Fig.18 is a side view of a fuel scraping portion provided by the present invention;

[0096] Fig.19 It is a schematic diagram of the implementation flow of the low-load stable combustion monitoring and control method for a coal-fired boiler provided by the present invention.

[0097] In the figure: 10-stable combustion execution module, 11-equipment base, 12-pretreatment seat, 121-pretreatment chamber, 122-discharge conduction seat, 13-fuel release pipe, 131-fuel injection port, 2-pretreatment mechanism, 21-pretreatment motor, 22-pretreatment drive unit, 221-driving gear, 222-driven gear, 223-linking gear ring, 224-driven gear ring, 23-pretreatment linkage unit, 231-pretreatment linkage rod, 232-first rotor, 233-second rotor, 234-pretreatment driven rod, 24-fuel refinement unit , 241-main refining roller, 242-auxiliary refining roller, 243-turbine refining roller, 244-turbine crushing teeth, 3-anti-accumulation component, 31-drive motor, 32-feeding auger, 33-fuel scraping part, 331-scraping mounting seat, 332-scraping rotating arm, 333-casing drive roller, 334-adsorbed material scraping roller, 34-anti-adsorption casing, 200-data acquisition module, 300-load prediction module, 310-data processing unit, 320-model definition unit, 330-prediction execution unit, 400-optimization control module. DETAILED DESCRIPTION

[0098] In order to enable those skilled in the art to better understand the technical solution of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0099] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present application. The terms used herein in the specification of the application are only for the purpose of describing specific embodiments and are not intended to limit the present application. The terms "including" and "having" in the specification and claims of the present application and the above-mentioned accompanying drawings and any variations thereof are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned accompanying drawings are used to distinguish different objects, rather than to describe a specific order.

[0100] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0101] When the existing method determines the combustion and stabilization index of a coal-fired boiler by the flame combustion state of the coal-fired boiler, the combustion and stabilization index is affected by the real-time load parameters and the boiler operation status data. A single flame combustion state cannot fully reflect the combustion process. In view of the above problems, we propose a low-load stable combustion monitoring and control device and method for a coal-fired boiler. In short, the control device consists of a data acquisition module 200, a load prediction module 300, an optimization control module 400, and a stable combustion execution module 10. When working, the data acquisition module 200 first obtains the operation status data and real-time load value of the coal-fired boiler, and then the load prediction module 300, combining the conditional risk value method and the long short-term memory neural network model, constructs a load prediction model. The load prediction model identifies the operation status data and the real-time load value, and outputs a load power prediction sequence. Then the optimization control module 400 calculates the stable combustion parameters of the coal-fired boiler based on the load power prediction sequence, and superimposes the stable combustion parameters of the coal-fired boiler with the load power prediction sequence to obtain a predicted stable combustion optimization instruction. Finally, the stable combustion execution module 10 responds to the predicted stable combustion optimization instruction and performs the stable combustion optimization work. In the embodiment of the present invention, a load prediction module 300 is provided. The load prediction module 300 constructs a load prediction model by combining the conditional risk value method and the long short-term memory neural network model, thereby effectively combining the real-time load parameters and the boiler operation status data, and realizing the accurate prediction of the load power, thereby ensuring the low-load stable combustion optimization operation of the coal-fired boiler, and overcoming the problem that when the existing method determines the combustion stability index of the coal-fired boiler through the flame combustion state of the coal-fired boiler, the combustion stability index is affected by the real-time load parameters and the boiler operation status data, and a single flame combustion state cannot fully reflect the combustion process.

[0102] The embodiment of the present invention provides a low-load stable combustion monitoring and control device for a coal-fired boiler. Figure 1 The structure diagram of the low-load stable combustion monitoring and control device for a coal-fired boiler is shown, and the low-load stable combustion monitoring and control device for a coal-fired boiler specifically includes:

[0103] The data acquisition module 200 is used to obtain the operating status data and real-time load value of the coal-fired boiler.

[0104] It should be noted that the data acquisition module 200 includes a flame monitor, a temperature sensor, a humidity sensor, a pressure sensor, a flow sensor, a vibration sensor, and an environmental monitoring sensor, and the operating status data includes but is not limited to furnace temperature, superheater temperature, steam pressure, feed water pressure, fuel flow, air flow, steam flow, flue gas composition, boiler vibration, motor current, voltage, and emission concentration data.

[0105] The load forecasting module 300 combines the conditional value at risk method and the long short-term memory neural network model to build a load forecasting model. The load forecasting model identifies the operating status data and the real-time load value and outputs a load power forecast sequence.

[0106] The optimization control module 400 is used to obtain the load power prediction sequence, calculate the coal-fired boiler stable combustion parameters based on the load power prediction sequence, and superimpose the coal-fired boiler stable combustion parameters with the load power prediction sequence to obtain the prediction stable combustion optimization instruction.

[0107] The combustion stabilization execution module 10 executes the combustion stabilization optimization work in response to the predicted combustion stabilization optimization instruction.

[0108] In this embodiment, the data acquisition module 200 is connected to the load prediction module 300 , the optimization control module 400 , and the stable combustion execution module 10 through communication, thereby realizing interactive transmission of data.

[0109] In the embodiment of the present invention, a load prediction module 300 is provided. The load prediction module 300 constructs a load prediction model by combining the conditional risk value method and the long short-term memory neural network model, thereby effectively combining the real-time load parameters and the boiler operation status data, and realizing the accurate prediction of the load power, thereby ensuring the low-load stable combustion optimization operation of the coal-fired boiler, and overcoming the problem that when the existing method determines the combustion stability index of the coal-fired boiler through the flame combustion state of the coal-fired boiler, the combustion stability index is affected by the real-time load parameters and the boiler operation status data, and a single flame combustion state cannot fully reflect the combustion process.

[0110] It should be noted that the load prediction module 300 includes:

[0111] The data processing unit 310 is used to obtain the operation status data and the real-time load value, and decompose the operation status data to obtain a state decomposition set.

[0112] The model definition unit 320 loads the state decomposition set and the real-time load value, trains a load forecasting model based on the conditional value at risk method and the long short-term memory neural network model based on the state decomposition set and the real-time load value, and outputs the load forecasting model.

[0113] The prediction execution unit 330, the load prediction model identifies the operation status data and the real-time load value, and outputs the load power prediction sequence.

[0114] In an embodiment of the present invention, a load prediction module 300 is provided. The load prediction module 300 is composed of a data processing unit 310, a model definition unit 320 and a prediction execution unit 330. The setting of the load prediction module 300 can accurately predict the load of the coal-fired boiler, thereby providing data support for the stable operation of the coal-fired boiler at low load.

[0115] The embodiment of the present invention provides a method for decomposing and processing running status data. Figure 2 The schematic diagram of the implementation flow of the method for decomposing and processing the running status data is shown. The method for decomposing and processing the running status data specifically includes:

[0116] Step S101, decomposing the running status data based on the empirical mode decomposition algorithm to obtain a random oscillation sequence.

[0117]

[0118] q represents the operation status data decomposition coefficient. In this embodiment, the operation status data decomposition coefficient can be 0.2-0.5, u i is the random oscillation sequence during decomposition, Q n represents the decomposition trend term, and x(t) represents the input operating status data.

[0119]

[0120] maxx(t) and minx(t) represent the local maximum and local minimum values ​​in the input running status data, respectively. Represents the local mean of the input operating status data.

[0121] Step S102, obtain a random oscillation sequence, introduce white noise with an amplitude of δ and an expansion factor α into the random oscillation sequence. In this embodiment, the amplitude is 0.1-0.2, and the expansion factor is 0.02-0.05, reconstruct the random oscillation sequence, and obtain a reconstructed oscillation sequence.

[0122] The reconstructed oscillation sequence output is expressed as:

[0123]

[0124] Step S103, calculating the overall mean of the reconstructed oscillation sequence, and taking the overall mean as the state decomposition set.

[0125] In an embodiment of the present invention, the operating status data is decomposed based on the empirical mode decomposition algorithm, and a random oscillation sequence is reconstructed. By decomposing the operating status data, the performance of the boiler under different operating conditions can be predicted, and the operating parameters can be adjusted accordingly to achieve optimal combustion efficiency.

[0126] The embodiment of the present invention provides a method for training a load forecasting model based on a conditional value at risk method and a long short-term memory neural network model based on a state decomposition set and real-time load values. Figure 3 The schematic diagram of the implementation flow of the load forecasting model method based on the state decomposition set and real-time load value training based on the conditional value at risk method and the long short-term memory neural network model is shown. The load forecasting model based on the state decomposition set and real-time load value training based on the conditional value at risk method and the long short-term memory neural network model specifically includes:

[0127] Step S201, taking the long short-term memory neural network model as the initial model, the initial model includes an input layer, a hidden layer, and an output layer, and the input layer introduces peak load balance constraints, load power balance constraints, stable combustion burner output constraints, and coal-fired boiler ramp rate constraints.

[0128] In this embodiment, the input layer is connected to the hidden layer, and the hidden layer is connected to the output layer. The number of neurons in the input layer is 32. The input layer introduces peak-shaving balance constraints, load power balance constraints, stable combustion burner output constraints, and coal-fired boiler climbing rate constraints, which can quickly capture the state decomposition set and real-time load value characteristic factors, and conditionally constrain the characteristic factors, thereby facilitating the construction of the load forecasting model.

[0129] Step S202: The conditional value at risk algorithm is introduced into the hidden layer to construct a low load objective function of the load forecasting model.

[0130] Among them, the low load objective function is defined as follows:

[0131]

[0132] y loa (t+1) is the load power forecast sequence for the next cycle, y loa (t) is the current load power, y loa (t-1) is the load power of the previous cycle, w y represents the peak load balance constraint, v trepresents the ramp rate constraint of the coal-fired boiler, w t represents the output constraint of the stable combustion burner, represents the rated output constraint of the stable combustion burner, Q r is the total work output of the stable combustion burner, g y is the inertia coefficient of the stable combustion burner.

[0133]

[0134] Step S203, obtaining a state decomposition set and a real-time load value, and extracting characteristic factors corresponding to the state decomposition set and the real-time load value at the input layer to obtain a characteristic factor set.

[0135] Step S204: The hidden layer uses the low-load objective function as an excitation function, assigns weights to the characteristic factor set, calculates the node output of the hidden layer, and obtains a load power prediction sequence.

[0136] Step S205: Use mean absolute error and root mean square error to verify the difference between the load power forecast sequence and the actual sequence.

[0137] Step S206, verifying the accuracy of the model. In this embodiment, the accuracy of the model can be set to 0.8-0.85.

[0138] Step S207: if the model meets the preset accuracy of mean absolute error and root mean square error, output the load power prediction sequence.

[0139] If it does not meet the accuracy of the model, return to step S203.

[0140] In an embodiment of the present invention, a load forecasting model is provided, which uses a long short-term memory neural network model as an initial model and introduces a conditional value at risk algorithm to complete the construction of the load forecasting model, thereby ensuring the prediction accuracy and robustness of the load forecasting model that combines the long short-term memory neural network model and the conditional value at risk algorithm.

[0141] The embodiment of the present invention provides a method for calculating stable combustion parameters of a coal-fired boiler based on a load power prediction sequence. Figure 4 The schematic diagram of the implementation flow of the method for calculating the stable combustion parameters of a coal-fired boiler based on a load power prediction sequence is shown. The method for calculating the stable combustion parameters of a coal-fired boiler based on a load power prediction sequence specifically includes:

[0142] Step S301, loading a load power prediction sequence, taking the difference between the load power prediction sequence and the output power of the coal-fired boiler in a stable combustion period tending to a set difference as a constraint, and calculating the stable combustion parameters of the coal-fired boiler.

[0143]

[0144] Among them, minΔyloa To set the difference constraint, is the load power forecast sequence y within the set period T loa (t+1), β is the stable combustion control command of the coal-fired boiler allocated by the stable combustion burner.

[0145] Step S302, obtaining the stable combustion parameters of the coal-fired boiler, correcting the stable combustion parameters of the coal-fired boiler based on the rolling optimization algorithm, and outputting the corrected stable combustion parameters of the coal-fired boiler.

[0146] The embodiment of the present invention provides a combustion stabilization execution module 10. Figure 5-Figure 7 The structural diagram of the stable combustion execution module 10 is shown, and the stable combustion execution module 10 specifically includes:

[0147] The equipment base 11 is fixedly mounted on the side wall of the coal-fired boiler through a support rod or a clamp.

[0148] A pretreatment seat 12 is fixedly mounted on the equipment base 11, a pretreatment chamber 121 is arranged in the pretreatment seat 12, a discharge guide seat 122 is provided at the bottom of the pretreatment seat 12, a pretreatment mechanism 2 is arranged in the pretreatment chamber 121, and the pretreatment mechanism 2 is used to pretreat the stable combustion fuel.

[0149] It should be noted that the pre-treatment seat 12 may be a rectangular seat or a round seat with a hollow interior, and the side wall or the bottom of the pre-treatment seat 12 is fixedly mounted on the equipment base 11 by fastening bolts or welding.

[0150] The fuel release pipe 13 is arranged on the equipment base 11, one end of the fuel release pipe 13 is fixedly connected to the discharge guide seat 122, the other end of the fuel release pipe 13 is provided with a fuel injection port 131, and an anti-deposition component 3 is arranged in the fuel release pipe 13.

[0151] In this embodiment, the inner wall of the fuel release tube 13 is polished, and the fuel release tube 13 can be a circular tube with a hollow interior.

[0152] The embodiment of the present invention provides a pre-processing mechanism 2, Figure 8-Figure 11 The schematic diagram of the structure of the pre-processing mechanism 2 is shown, and the pre-processing mechanism 2 specifically comprises:

[0153] The pre-processing motor 21 is fixedly mounted on the equipment base 11 . The pre-processing motor 21 is fixedly mounted on the equipment base 11 by a clamp or fastening bolts. The pre-processing motor 21 may be a servo motor.

[0154] A pre-processing driving unit 22 is fixedly connected to the output end of the pre-processing motor 21 .

[0155] The pre-processing linkage part 23 is disposed on the pre-processing seat 12 , and the pre-processing linkage part 23 is connected to the pre-processing driving part 22 .

[0156] The fuel refinement section 24, such as Fig.12 As shown, the fuel refinement section 24 is arranged in the pretreatment chamber 121, and the fuel refinement section 24 includes a main refinement roller 241, an auxiliary refinement roller 242 and a spoiler refinement roller 243. The main refinement roller 241 and the auxiliary refinement roller 242 are respectively connected to the pretreatment driving section 22, the spoiler refinement roller 243 is connected to the pretreatment linkage section 23, and the side wall of the spoiler refinement roller 243 is fixedly installed with a spoiler crushing tooth 244.

[0157] It should be noted that one end of the main refining roller 241 and the auxiliary refining roller 242 are rotatably connected to the side wall of the pretreatment seat 12 through bearings or rollers. The main refining roller 241 and the auxiliary refining roller 242 are symmetrically arranged, and the shape can be a truncated cone or cylindrical roller structure, while the spoiler refining roller 243 can be an arc or semicircular structure. The spoiler refining roller 243 is arranged between the main refining roller 241 and the auxiliary refining roller 242. The spoiler breaking teeth 244 can be a pointed cone or prismatic structure, and the spoiler breaking teeth 244 are arranged on the inner and outer sides of the spoiler refining roller 243.

[0158] In the embodiment of the present invention, a pretreatment mechanism 2 is provided, and the pretreatment mechanism 2 is composed of a pretreatment drive unit 22, a pretreatment linkage unit 23, and a fuel refinement unit 24. The pretreatment drive unit 22, the pretreatment linkage unit 23, and the fuel refinement unit 24 cooperate with each other to achieve sufficient refinement of the fuel pulverized coal in the pretreatment chamber 121. On the one hand, it ensures the uniform injection of the fuel from the fuel injection port 131, and on the other hand, it is also conducive to the stable combustion and low-load stable operation of the coal-fired boiler.

[0159] In this embodiment, the pre-processing driving unit 22 includes:

[0160] The driving gear 221 is arranged on one side of the pre-processing motor 21 , and is fixedly connected to the output end of the pre-processing motor 21 . The side of the driving gear 221 away from the pre-processing motor 21 is fixedly connected to the end of the main thinning roller 241 .

[0161] The driven gear 222 is rotatably mounted on the side wall of the pre-treatment seat 12 , and the driven gear 222 is meshed with the driving gear 221 for transmission, and one side of the driven gear 222 is fixedly connected to the end of the auxiliary thinning roller 242 .

[0162] A linkage ring gear 223 , which is fixedly mounted on the outer side wall of the driven gear 222 , and a driven ring gear 224 meshing with the linkage ring gear 223 , which is fixedly connected to the pre-treatment linkage part 23 .

[0163] In this embodiment, the driving gear 221 and the driven gear 222 are rotatably connected to the side wall of the pretreatment seat 12 through bearings, one side of the driving gear 221 is fixedly connected to the output end of the pretreatment motor 21 by plugging or riveting, and the linkage ring gear 223 is fixedly installed on the outer wall of the driven gear 222 by plugging or riveting, and the driven ring gear 224 is rotatably connected to the side wall of the pretreatment seat 12 through a bearing or a limit sleeve.

[0164] It should be noted that the pre-processing linkage unit 23 includes:

[0165] A pre-treatment linkage rod 231 , one end of which is fixedly connected to the driven gear ring 224 , and the pre-treatment linkage rod 231 is rotatably mounted on the side wall of the pre-treatment seat 12 .

[0166] The first rotating wheel 232 is fixedly mounted on the top of the pre-processing linkage rod 231 .

[0167] The second rotating wheel 233 is disposed at one side of the first rotating wheel 232 , and the second rotating wheel 233 is rotatably connected to the first rotating wheel 232 via a transmission belt.

[0168] A pre-treatment driven rod 234 , one end of which is fixedly connected to the second rotating wheel 233 , and the other end of which passes through the top wall of the pre-treatment seat 12 and is fixedly connected to the turbulence refinement roller 243 .

[0169] It should be noted that the upper and lower ends of the pretreatment linkage rod 231 are fixedly connected to the first rotating wheel 232 and the driven gear ring 224 respectively by plugging, and the second rotating wheel 233 is also fixedly connected to the pretreatment driven rod 234 by plugging, and the end of the pretreatment driven rod 234 away from the second rotating wheel 233 is welded or riveted to the upper end of the spoiler refinement roller 243.

[0170] During operation, the blocky coal powder enters the pretreatment chamber 121, and then the pretreatment motor 21 is turned on. The start of the pretreatment motor 21 can drive the driving gear 221 and the driven gear 222 to rotate. The driving gear 221 and the driven gear 222 respectively drive the main refining roller 241 and the auxiliary refining roller 242 to rotate, thereby achieving full crushing of the fuel coal powder in the pretreatment chamber 121. At the same time, when the driven gear 222 rotates, it can drive the linkage gear ring 223 and the driven gear ring 224 to rotate. When the driven gear ring 224 rotates, it can drive the pretreatment linkage rod 231 to rotate. When the pretreatment linkage rod 231 rotates, it can drive the first runner 232, the second runner 233 and the pretreatment driven rod 234 to rotate. The rotation of the pretreatment driven rod 234 can drive the turbulence refining roller 243 to rotate, thereby further achieving full crushing of the fuel coal powder.

[0171] In a further preferred embodiment of the present invention, Figure 13-Figure 16 As shown, the anti-deposit component 3 includes:

[0172] The driving motor 31 is fixedly mounted on the device base 11 . The driving motor 31 is fixedly mounted on the device base 11 by means of a clamping ring or fastening bolts.

[0173] The material feeding auger 32 is disposed in the fuel release pipe 13 , and one end of the material feeding auger 32 is fixedly connected to the output end of the driving motor 31 .

[0174] The anti-adsorption sleeve 34 is rotatably mounted in the fuel release pipe 13 , and the anti-adsorption sleeve 34 is rotatably mounted on the end wall of the fuel release pipe 13 .

[0175] It should be noted that the anti-adsorption sleeve 34 may be a circular tube or a circular ring with a hollow interior, and the inner wall of the anti-adsorption sleeve 34 is polished.

[0176] The fuel scraping portion 33 is disposed in the fuel release pipe 13 , and the fuel scraping portion 33 is fixedly connected to the end of the feeding auger 32 .

[0177] Among them, Figure 17-Figure 18 As shown, the fuel scraping portion 33 includes:

[0178] A scraper linkage rod is fixedly connected to the end of the feeding auger 32 .

[0179] At least one set of scraper mounting seats 331, wherein the scraper mounting seats 331 are fixedly sleeved on the scraper linkage rod.

[0180] A scraping rotating arm 332 , the scraping rotating arm 332 is fixedly mounted on the scraping mounting seat 331 , and a sleeve driving roller 333 is fixedly mounted on the scraping rotating arm 332 , the sleeve driving roller 333 is used to drive the anti-adsorption sleeve 34 .

[0181] The adsorbed matter scraping roller 334 is rotatably mounted on the sleeve driving roller 333 .

[0182] In this embodiment, the scraper linkage rod and the conveying auger are fixedly connected by mortise and tenon, plug-in or riveting. The scraper mounting seat 331 is arranged on the scraper linkage rod by plug-in or welding. The scraper rotating arm 332 is circumferentially arranged on the scraper mounting seat 331. The sleeve drive roller 333 is fixedly installed on the end of the scraper rotating arm 332 by welding or riveting. The outer wall of the sleeve drive roller 333 is provided with a rectangular and spiral anti-slip groove structure.

[0183] On the other hand, the embodiment of the present invention also provides a low-load stable combustion monitoring and control method for a coal-fired boiler. Fig.19 The schematic diagram of the implementation process of the low-load stable combustion monitoring and control method of the coal-fired boiler is shown. The low-load stable combustion monitoring and control method of the coal-fired boiler specifically includes:

[0184] Step S10, obtaining the operating status data and real-time load value of the coal-fired boiler.

[0185] Step S20, a load forecasting model is constructed by combining the conditional value at risk method and the long short-term memory neural network model. The load forecasting model identifies the operating status data and the real-time load value, and outputs a load power forecasting sequence.

[0186] Step S30, obtaining a load power prediction sequence, calculating the stable combustion parameters of the coal-fired boiler based on the load power prediction sequence, and superimposing the stable combustion parameters of the coal-fired boiler with the load power prediction sequence to obtain a prediction stable combustion optimization instruction.

[0187] Step S40, in response to the predicted stable combustion optimization instruction, executing the stable combustion optimization work.

[0188] On the other hand, the present invention also provides a computer device, which includes a display screen, a memory, a processor and a computer program, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the coal-fired boiler low-load stable combustion monitoring and control method.

[0189] It is understandable that, in the preferred embodiment provided by the present invention, the computer device may also be a notebook computer, a personal digital assistant (PDA), a mobile phone or other device capable of communication.

[0190] Exemplarily, the computer program may be divided into one or more modules, one or more modules are stored in a memory and executed by a processor to complete the present invention. One or more modules may be a series of computer program instruction segments capable of completing specific functions, and the instruction segments are used to describe the execution process of the computer program in a terminal device. For example, the above-mentioned computer program may be divided into units or modules of the coal-fired boiler low-load stable combustion monitoring and control system provided by the above-mentioned various system embodiments.

[0191] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the terminal device, and uses various interfaces and lines to connect various parts of the entire user terminal.

[0192] In summary, the present invention provides a low-load stable combustion monitoring and control device and method for a coal-fired boiler. In an embodiment of the present invention, a load prediction module 300 is provided. The load prediction module 300 constructs a load prediction model by combining the conditional risk value method and the long short-term memory neural network model, thereby effectively combining real-time load parameters and boiler operation status data, and realizing accurate prediction of load power, thereby ensuring the low-load stable combustion optimization operation of the coal-fired boiler, and overcoming the problem that when the existing method determines the combustion and stable combustion index of the coal-fired boiler through the flame combustion state of the coal-fired boiler, the combustion and stable combustion index is affected by the real-time load parameters and the boiler operation status data, and a single flame combustion state cannot fully reflect the combustion process.

[0193] It should be noted that, for the above-mentioned embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described order of actions, because according to the present invention, some steps may be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0194] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the scope of protection of the invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on these embodiments, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in this field can still combine, add, delete or make other adjustments to the features in the various embodiments of the present invention according to the circumstances without conflict, without making creative work, so as to obtain different other technical solutions that do not deviate from the concept of the present invention in essence, and these technical solutions also belong to the scope of protection of the present invention.

Claims

1. A low-load stable combustion monitoring and control device for a coal-fired boiler, characterized in that: The low-load stable combustion monitoring and control device for a coal-fired boiler comprises: Data acquisition module, used to obtain the operating status data and real-time load value of the coal-fired boiler; The load forecasting module combines the conditional value at risk method and the long short-term memory neural network model to build a load forecasting model. The load forecasting model identifies the operating status data and the real-time load value and outputs the load power forecast sequence. An optimization control module is used to obtain a load power prediction sequence, calculate the stable combustion parameters of the coal-fired boiler based on the load power prediction sequence, and superimpose the stable combustion parameters of the coal-fired boiler with the load power prediction sequence to obtain a prediction stable combustion optimization instruction; The combustion stabilization execution module executes the combustion stabilization optimization work in response to the predicted combustion stabilization optimization instruction.

2. The low-load stable combustion monitoring and control device for a coal-fired boiler according to claim 1, characterized in that: The load forecasting module comprises: A data processing unit is used to obtain operation status data and real-time load values, and decompose the operation status data to obtain a state decomposition set; The method for decomposing and processing the running status data specifically includes: Based on the empirical mode decomposition algorithm, the operating status data is decomposed to obtain a random oscillation sequence; q represents the operating status data decomposition coefficient, u i is the random oscillation sequence during decomposition, Q n represents the decomposition trend term, and x(t) represents the input operation status data; maxx(t) and minx(t) represent the local maximum and local minimum values ​​in the input running status data, respectively. Represents the local mean of the input operating status data; Obtain a random oscillation sequence, introduce white noise with an amplitude of δ and an expansion factor α into the random oscillation sequence, reconstruct the random oscillation sequence, and obtain a reconstructed oscillation sequence; The reconstructed oscillation sequence output is expressed as: Calculate the overall mean of the reconstructed oscillation sequence and use the overall mean as the state decomposition set; A model definition unit is used to load a state decomposition set and a real-time load value, and to train a load forecasting model based on a conditional value at risk method and a long short-term memory neural network model based on the state decomposition set and the real-time load value, and output the load forecasting model; The prediction execution unit, the load prediction model identifies the operating status data and the real-time load value, and outputs the load power prediction sequence.

3. The low-load stable combustion monitoring and control device for a coal-fired boiler according to claim 2, characterized in that: The load forecasting model based on the state decomposition set and real-time load value training based on the conditional value at risk method and the long short-term memory neural network model specifically includes: The long short-term memory neural network model is used as the initial model. The initial model includes an input layer, a hidden layer, and an output layer. The input layer introduces peak load balance constraints, load power balance constraints, stable combustion burner output constraints, and coal-fired boiler ramp rate constraints. The conditional risk value algorithm is introduced into the hidden layer to construct the low load objective function of the load forecasting model; Among them, the low load objective function is defined as follows: y loa (t+1) is the load power forecast sequence for the next cycle, y loa (t) is the current load power, y loa (t-1) is the load power of the previous cycle, w y represents the peak load balance constraint, v t represents the ramp rate constraint of the coal-fired boiler, w t represents the output constraint of the stable combustion burner, represents the rated output constraint of the stable combustion burner, Q r is the total work output of the stable combustion burner, g y is the inertia coefficient of the stable combustion burner; Obtain a state decomposition set and a real-time load value, and extract characteristic factors corresponding to the state decomposition set and the real-time load value at the input layer to obtain a characteristic factor set; The hidden layer uses the low-load objective function as the excitation function, assigns weights to the characteristic factor set, calculates the node output of the hidden layer, and obtains the load power prediction sequence; The mean absolute error and root mean square error are used to verify the difference between the load power forecast sequence and the actual sequence, and to verify the accuracy of the model. If the model meets the preset accuracy of the mean absolute error and root mean square error, the load power forecast sequence is output.

4. The low-load stable combustion monitoring and control device for a coal-fired boiler according to claim 1, characterized in that: The method for calculating the stable combustion parameters of a coal-fired boiler based on a load power prediction sequence specifically includes: The load power prediction sequence is loaded, and the stable combustion parameters of the coal-fired boiler are calculated by taking the difference between the load power prediction sequence and the output power of the coal-fired boiler in the stable combustion period as a constraint to tend to the set difference; Among them, minΔy loa To set the difference constraint, is the load power forecast sequence y within the set period T loa (t+1), β is the stable combustion control command of the coal-fired boiler assigned by the stable combustion burner; Obtain stable combustion parameters of the coal-fired boiler, correct the stable combustion parameters of the coal-fired boiler based on a rolling optimization algorithm, and output the corrected stable combustion parameters of the coal-fired boiler.

5. The low-load stable combustion monitoring and control device for a coal-fired boiler according to any one of claims 2 to 4, characterized in that: The combustion stabilization execution module comprises: Equipment base; A pretreatment seat fixedly mounted on the base of the equipment, wherein a pretreatment chamber is arranged in the pretreatment seat, a discharge guide seat is provided at the bottom of the pretreatment seat, and a pretreatment mechanism is arranged in the pretreatment chamber, and the pretreatment mechanism is used to pretreatment the stable combustion fuel; A fuel release pipe is arranged on the equipment base, one end of the fuel release pipe is fixedly connected to the discharge guide seat, the other end of the fuel release pipe is provided with a fuel injection port, and an anti-deposition component is arranged in the fuel release pipe.

6. The low-load stable combustion monitoring and control device for a coal-fired boiler according to claim 5, characterized in that: The pre-processing mechanism comprises: A pre-processing motor, wherein the pre-processing motor is fixedly mounted on the equipment base; A pre-processing drive unit fixedly connected to an output end of the pre-processing motor; A pre-processing linkage part, which is arranged on the pre-processing seat and connected to the pre-processing driving part; The fuel refinement part is arranged in the pretreatment chamber, and the fuel refinement part includes a main refinement roller, an auxiliary refinement roller and a turbulent refinement roller. The main refinement roller and the auxiliary refinement roller are respectively connected to the pretreatment drive part, the turbulent refinement roller is connected to the pretreatment linkage part, and the side wall of the turbulent refinement roller is fixedly installed with turbulent crushing teeth.

7. The low-load stable combustion monitoring and control device for a coal-fired boiler according to claim 6, characterized in that: The pre-processing driving unit comprises: A driving gear, which is arranged on one side of the pre-processing motor, is fixedly connected to the output end of the pre-processing motor, and a side of the driving gear away from the pre-processing motor is fixedly connected to the end of the main refining roller; A driven gear, the driven gear is rotatably mounted on the side wall of the pre-treatment seat, the driven gear is meshed with the driving gear for transmission, and one side of the driven gear is fixedly connected to the end of the auxiliary thinning roller; A linkage gear ring, the linkage gear ring is fixedly mounted on the outer side wall of the driven gear, and a driven gear ring meshing with the linkage gear ring for transmission, the driven gear ring is fixedly connected to the pre-treatment linkage part.

8. The low-load stable combustion monitoring and control device for a coal-fired boiler according to claim 7, characterized in that: The pre-processing linkage unit comprises: A pretreatment linkage rod, one end of which is fixedly connected to the driven gear ring, and the pretreatment linkage rod is rotatably mounted on the side wall of the pretreatment seat; A first rotating wheel, wherein the first rotating wheel is fixedly mounted on the top of the pre-processing linkage rod; A second rotating wheel, the second rotating wheel is arranged on one side of the first rotating wheel, and the second rotating wheel is rotatably connected to the first rotating wheel through a transmission belt; A pretreatment driven rod, one end of which is fixedly connected to the second rotating wheel, and the other end of which passes through the top wall of the pretreatment seat and is fixedly connected to the turbulence refinement roller.

9. The low-load stable combustion monitoring and control device for a coal-fired boiler according to claim 5, characterized in that: The anti-accumulation component comprises: A driving motor, wherein the driving motor is fixedly mounted on a device base; A material conveying auger, wherein the material conveying auger is arranged in the fuel release pipe, and one end of the material conveying auger is fixedly connected to the output end of the driving motor; an anti-adsorption sleeve rotatably mounted in the fuel release pipe, wherein the anti-adsorption sleeve is rotatably mounted on an end wall of the fuel release pipe; A fuel scraping part, which is arranged in the fuel release pipe and is fixedly connected to the end of the feed auger; Wherein, the fuel scraping part comprises: A scraper linkage rod, the scraper linkage rod is fixedly connected to the end of the feeding auger; At least one set of scraper mounting seats, wherein the scraper mounting seats are fixedly sleeved on the scraper linkage rod; A scraping rotating arm, the scraping rotating arm is fixedly mounted on the scraping mounting seat, and a sleeve driving roller is fixedly mounted on the scraping rotating arm, the sleeve driving roller is used to drive the anti-adsorption sleeve; An adsorbent scraping roller is rotatably mounted on the sleeve driving roller.

10. A method for monitoring and controlling low-load stable combustion of a coal-fired boiler, implemented by using the low-load stable combustion monitoring and control device for a coal-fired boiler as claimed in any one of claims 1 to 9, characterized in that: The low-load stable combustion monitoring and control method of a coal-fired boiler specifically comprises: Obtain the operating status data and real-time load value of coal-fired boilers; The load forecasting model is constructed by combining the conditional risk value method and the long short-term memory neural network model. The load forecasting model identifies the operating status data and the real-time load value and outputs the load power forecast sequence. Obtain a load power forecast sequence, calculate the stable combustion parameters of the coal-fired boiler based on the load power forecast sequence, and superimpose the stable combustion parameters of the coal-fired boiler with the load power forecast sequence to obtain a forecast stable combustion optimization instruction; In response to the predicted combustion stabilization optimization instruction, a combustion stabilization optimization task is performed.

Images