Automatic boiler soot blowing system based on three-dimensional finite element neural network

Abstract

The invention relates to the technical field of automatic control, in particular to an automatic boiler soot blowing system based on a three-dimensional finite element neural network, which comprises a data acquisition module, an abnormal region delimiting module, a soot deposition region boundary determining module, a heat efficiency loss predicting module, a soot deposition core region judging module, a soot blowing control module and a management database, according to the system, an ash deposition area is screened through temperature, pressure and smoke flow data of a boiler heating surface, the boundary contour of the ash deposition area is determined, an ash deposition core area is obtained through calculation of ash deposition concentration, thickness and heat efficiency loss and comparison and screening with set threshold values, and an ash blowing path of the ash deposition core area is generated. Whether the soot blowing problem exists or not is judged according to the temperature data before and after soot blowing, the angle, needing to be adjusted, of the blowing head of the soot blower is calculated, it is ensured that a soot deposition core area is comprehensively and effectively blown, the soot blowing effect is improved, and soot deposition residues are reduced.

Description

Technical Field

[0001] The present invention relates to the field of automatic control technology, in particular to a boiler automatic soot blowing system based on a three-dimensional finite element neural network. Background Art

[0002] In modern industrial production, boilers, as core equipment for energy conversion, are widely used in multiple fields such as electricity, chemical industry, and heating. With the continuous expansion of industrial scale and the increasing demand for energy efficiency, the safe and stable operation and efficient heat transfer performance of boilers are crucial. However, the heating surface of boilers is prone to dust accumulation during long-term operation. Dust accumulation not only hinders heat transfer, reduces boiler thermal efficiency, and increases energy consumption, but can also cause local overheating and corrosion, seriously threatening the safe operation of the boiler. Therefore, timely and effective removal of dust from the boiler heating surface has become a key link in ensuring the efficient and safe operation of the boiler.

[0003] Traditional manual sootblowing is not only inefficient and labor-intensive, but also difficult to accurately control the timing and intensity of sootblowing, which can easily lead to problems such as incomplete or excessive sootblowing. With the continuous development of automation technology, automatic sootblowing systems are gradually being applied to the boiler industry. However, existing automatic sootblowing systems still have many shortcomings in soot accumulation detection and sootblowing strategy formulation, and cannot meet the high-quality boiler operation requirements of modern industry.

[0004] For example, the existing Chinese patent application number 202510255643.9 discloses an intelligent sootblowing method for the convection heating surface of a boiler. This scheme determines the sootblowing sequence that affects the heat exchange of the convection heating surface, determines the boundary conditions, calculates the maximum heat absorption and actual heat absorption under the boundary, and then calculates the real-time pollution rate. Based on the growth rate of the pollution rate, different sequences of sootblower are triggered to achieve targeted sootblowing.

[0005] However, this solution has the following shortcomings: First, this method mainly relies on the heat absorption of the convective heating surface to calculate the real-time pollution rate and determine whether soot blowing is needed. However, there are many factors that affect the heat absorption of the heating surface, and soot accumulation is only one of them. Fluctuations in flue gas flow, changes in combustion conditions, and abnormal steam parameters may all lead to changes in heat absorption, which in turn may cause deviations in the pollution rate calculation. Once interference factors occur, the system may misjudge the degree of soot accumulation and fail to accurately find the location of soot accumulation and serious areas, resulting in a lack of targeted soot blowing and difficulty in effectively removing soot accumulation.

[0006] Second, the solution lacks real-time feedback and optimization of the sootblowing effect. Although the DCS system monitors the sootblower operating status in real time, this method has not established a complete real-time feedback mechanism for the sootblowing effect. After the sootblowing is completed, the system does not conduct in-depth analysis and evaluation of the actual changes in the contamination rate of the heated surface, and cannot promptly determine whether the sootblowing has achieved the expected effect. If the sootblowing effect is not good, the system cannot automatically adjust the sootblowing parameters or re-plan the sootblowing sequence, making it difficult to achieve continuous optimization of the sootblowing effect. Summary of the Invention

[0007] In order to overcome the shortcomings of the background technology, an embodiment of the present invention provides a boiler automatic sootblowing system based on a three-dimensional finite element neural network, which can effectively solve the problems involved in the above background technology.

[0008] The purpose of the present invention can be achieved through the following technical solutions: The present invention provides a boiler automatic soot blowing system based on a three-dimensional finite element neural network, including: a data acquisition module for respectively acquiring the temperature, pressure and flue gas flow data of the boiler heating surface.

[0009] The abnormal area delineation module is used to screen abnormal areas according to the temperature, pressure and flue gas flow data of the boiler heating surface, and select the overlapping parts of the abnormal areas to be marked as dust accumulation areas.

[0010] The dust accumulation area boundary determination module is used to search for edge points in the dust accumulation area to form a boundary contour of the dust accumulation area.

[0011] The thermal efficiency loss prediction module is used to calculate the ash accumulation thickness and concentration, build a thermal efficiency loss prediction model through simulation data, and calculate the thermal efficiency loss prediction value.

[0012] The dust accumulation core area determination module is used to compare the dust accumulation concentration, thickness and thermal efficiency loss with the set threshold value to screen out the dust accumulation core area.

[0013] The sootblowing control module is used to generate the sootblowing path in the soot accumulation core area, determine whether there is a sootblowing problem based on the temperature data before and after sootblowing, and calculate the angle at which the sootblower blowing head needs to be adjusted.

[0014] Management database, used to store measurement data, abnormal area data, dust accumulation area data, thermal efficiency loss prediction model data and soot blowing control data.

[0015] Preferably, the specific analysis method of the data acquisition module is: select a number of equally spaced measurement points on the surface of the boiler heating surface, obtain the temperature and pressure of each measurement point through a temperature sensor and a pressure sensor respectively, and at the same time divide the time points into equal intervals, and detect the flue gas flow at each measurement point at each time point to obtain the flow at each measurement point at each time point.

[0016] Preferably, the specific analysis method of the abnormal area delineation module is: sorting the collected temperature data in sequence according to the position of the measurement point, taking adjacent measurement points as a group, calculating the temperature difference at each group of adjacent measurement points, and then dividing it by the distance between each group of adjacent measurement points to obtain the temperature gradient between each group of adjacent measurement points, setting the temperature gradient standard value, screening out each group of adjacent measurement points whose temperature gradient is greater than the temperature gradient standard value, and delineating the abnormal temperature change area according to their corresponding positions.

[0017] Calculate the pressure difference at each group of adjacent measurement points, and obtain the pressure gradient between each group of adjacent measurement points by dividing the pressure difference by the distance between each group of adjacent measurement points. Set a pressure gradient standard value, and screen out each group of adjacent measurement points with a pressure gradient greater than the pressure gradient standard value. Based on their corresponding positions, delineate the area of ​​abnormal pressure change.

[0018] Arrange the flow data in chronological order, calculate the flow change at each adjacent time point at each measuring point, and obtain the flow change rate at each adjacent time point at each measuring point by dividing it by the time interval. Subtract the flow change rate from the average flow change rate at each adjacent time point to obtain the flow change rate difference at each measuring point. Set a difference standard, screen out the measuring points with a flow change rate difference greater than the difference standard, and delineate the flow abnormality area based on their corresponding positions.

[0019] Read the abnormal temperature change area, abnormal pressure change area, and abnormal flow area respectively, and select the overlapping area to mark as the dust accumulation area.

[0020] Preferably, the specific analysis method of the dust accumulation area boundary determination module is: the first step is to select a point from the dust accumulation area as the starting point, starting from the starting point, searching for adjacent points in the set direction, and obtaining the temperature, pressure, and flow data of the starting point and the adjacent point respectively, and calculating the temperature change, pressure change, and flow change based on the obtained temperature, pressure, and flow data of the starting point and the adjacent point.

[0021] In the second step, each change is compared with the corresponding preset threshold. If the change of any parameter between adjacent points exceeds the corresponding preset threshold, the adjacent point is identified as an edge point and the coordinates of the point are recorded. If the change of all parameters does not exceed the threshold, the adjacent point is judged not to be an edge point.

[0022] In the third step, after determining that the adjacent point is not an edge point, continue to search for the adjacent points of the adjacent point according to the above operation until a new edge point is found, and use the newly identified edge point as the current point, repeating the above operation until the search returns to the starting point.

[0023] The fourth step is to connect all edge points identified and recorded during the search process in a certain order to form the boundary outline of the dust accumulation area.

[0024] Preferably, the specific analysis method of the dust accumulation thickness and dust accumulation concentration is: construct a regular grid with the boundary outline of the dust accumulation area as the range, select the grid intersection as each detection position, apply coupling agent at each selected detection position, transmit ultrasonic pulses through the ultrasonic transducer, vertically incident on the dust layer, and record the time from emission to reception of the reflected wave, calculate the average value through multiple measurements, and obtain the propagation time of the ultrasonic wave in the dust at each detection position, and at the same time obtain the reflection intensity of the ultrasonic wave reflected back to the ultrasonic transducer at each detection position from the equipment.

[0025] The dust thickness at each detection position is calculated based on the propagation time and propagation speed of the ultrasonic wave in the dust at each detection position, and the dust concentration at each detection position is calculated based on the calculated dust thickness at each detection position and the reflection intensity of the ultrasonic wave reflected back to the ultrasonic transducer at each detection position.

[0026] Preferably, the specific analysis method of the thermal efficiency loss prediction module is: setting a series of dust accumulation thicknesses and dust accumulation concentrations at fixed intervals, corresponding them one by one, and obtaining each dust accumulation thickness and concentration combination; for each dust accumulation thickness and concentration combination, constructing a geometric model of the heat transfer system, and dividing the grid, setting boundary conditions and initial conditions for it respectively; obtaining the heat flux density of the heat transfer surface according to Fourier's law, integrating the heat transfer surface, and calculating the heat transfer rate according to the geometric shape and heat flux density of the heat transfer surface.

[0027] Under the condition of no dust accumulation, the heat transfer rate without dust accumulation is simulated and calculated, and the thermal efficiency loss under the current dust accumulation parameters is calculated according to the definition formula of thermal efficiency loss. In this way, each set of dust accumulation parameters is simulated and calculated to obtain the thermal efficiency loss under each combination of dust thickness and concentration.

[0028] A polynomial regression equation is constructed based on the thermal efficiency loss under various combinations of dust thickness and concentration obtained by simulation. The values ​​of various coefficients are calculated through the regression algorithm to determine the thermal efficiency loss prediction model, and the thermal efficiency loss prediction value is calculated accordingly.

[0029] The dust accumulation thickness and dust accumulation concentration at each detection position are substituted into the thermal efficiency loss prediction model to obtain the thermal efficiency loss prediction value at each detection position.

[0030] Preferably, the specific analysis method of the dust accumulation core area determination module is: the first step is to set thresholds for dust concentration, thickness and thermal efficiency loss respectively, read the dust thickness, dust concentration and thermal efficiency loss prediction value of each detection position, and compare them with the corresponding thresholds respectively. If the dust thickness, dust concentration and thermal efficiency loss prediction value of a certain detection position are greater than or equal to the corresponding threshold, the area where the detection position is located is preliminarily determined to be the dust accumulation core detection point.

[0031] In the second step, with the dust core detection point as the center, the dust core detection point is judged for each vertex of its four adjacent grids. If among the four vertices of a certain adjacent grid, there are dust thicknesses, dust concentrations, and thermal efficiency loss prediction values ​​greater than or equal to two vertices that are greater than or equal to the corresponding thresholds, then the grid is judged to belong to the dust core detection area. In this way, the determined dust core detection areas are connected to obtain the dust core area.

[0032] Preferably, the specific analysis method of the soot blowing path in the soot accumulation core area is: obtain the position and range of the soot accumulation core area, divide it into several sub-areas according to the shape of the soot accumulation core area, determine the blowing order of each sub-area according to the set logic, plan the soot blowing path according to the shape of each sub-area, and connect the soot blowing paths planned in each sub-area to form a continuous soot blowing path, so as to control the soot blower to perform soot blowing.

[0033] Preferably, the specific analysis method for determining whether there is a soot blowing problem is: select several soot blowing detection points in the core area of ​​boiler soot accumulation, obtain the temperature of each soot blowing detection point before and after soot blowing through a temperature sensor, take the difference and divide it by the soot blowing time to obtain the temperature change rate of each soot blowing detection point; if the temperature change rate of a soot blowing detection point is positive and reaches the set threshold, it is determined that the soot accumulation is reduced and the heat transfer effect is improved; if the temperature change rate of a soot blowing detection point does not reach the set threshold, it is determined that there is a soot blowing problem and the soot blowing angle needs to be adjusted.

[0034] Preferably, the specific analysis method for the angle that the sootblower blowing head needs to be adjusted is: obtain the sub-area to which the sootblowing detection point where the sootblowing problem exists belongs, obtain the width of the soot accumulation area of ​​the sub-area, measure the width of the soot accumulation area that can be covered by the current sootblowing angle, and make a difference to obtain the width of the soot accumulation area not covered by the current sootblowing angle. Combined with the sootblower blowing coverage width, the angle that the sootblower blowing head needs to be adjusted is calculated, and the sootblower blowing head is adjusted accordingly.

[0035] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention screens the ash accumulation area through the temperature, pressure and flue gas flow data of the boiler heating surface, determines the boundary contour of the ash accumulation area, and can accurately screen out the abnormal temperature, pressure and flow areas, and then determine the ash accumulation area and its boundary contour, avoiding the ambiguity and uncertainty of judging the ash accumulation location based on experience alone.

[0036] 2. The present invention calculates the ash concentration, thickness and thermal efficiency loss, and compares them with the set threshold value to screen out the ash core area, which can clearly identify the area that needs to be treated, avoids blind soot blowing on the entire heating surface, improves the pertinence and effectiveness of soot blowing, reduces unnecessary soot blowing operations, and reduces the wear and energy consumption of the soot blower.

[0037] 3. The present invention generates a soot blowing path for the core area of ​​soot accumulation, determines whether there is a soot blowing problem based on the temperature data before and after soot blowing, and calculates the angle at which the soot blower head needs to be adjusted, thereby ensuring that the core area of ​​soot accumulation is comprehensively and effectively blown, thereby improving the soot blowing effect and reducing soot residue. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0039] Figure 1 This is a system module connection diagram of the present invention.

[0040] Figure 2 for Figure 1 Flowchart of the module for determining the boundary of the dust accumulation area.

[0041] Figure 3 for Figure 1 Flowchart of the medium dust accumulation core area determination module. DETAILED DESCRIPTION

[0042] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0043] See also Figure 1As shown, a boiler automatic soot blowing system based on a three-dimensional finite element neural network includes a data acquisition module, an abnormal area demarcation module, a soot accumulation area boundary determination module, a thermal efficiency loss prediction module, a soot accumulation core area determination module, a soot blowing control module, and a management database.

[0044] The management database is connected to the dust accumulation area boundary determination module, the thermal efficiency loss prediction module, the dust accumulation core area judgment module, and the soot blowing control module; the abnormal area demarcation module is connected to the data acquisition module and the dust accumulation area boundary determination module; the dust accumulation area boundary determination module is connected to the thermal efficiency loss prediction module and the dust accumulation core area judgment module; and the thermal efficiency loss prediction module is connected to the dust accumulation core area judgment module.

[0045] The data acquisition module is used to obtain the temperature, pressure and flue gas flow data of the boiler heating surface respectively.

[0046] The specific analysis method of the data acquisition module is as follows: a number of equally spaced measurement points are selected on the surface of the boiler heating surface, and the temperature and pressure of each measurement point are obtained by temperature sensors and pressure sensors respectively. At the same time, the time points are divided into equal intervals, and the flue gas flow rate at each measurement point at each time point is detected respectively to obtain the flow rate of each measurement point at each time point; the temperature and pressure of each measurement point are obtained in real time by using temperature sensors and pressure sensors. Combined with the detection of flue gas flow rate, the operating status of the boiler at different time points can be accurately monitored, abnormal changes in temperature, pressure and flow rate can be discovered in time, and data support can be provided for the safe and stable operation of the boiler.

[0047] The abnormal area delineation module is used to screen abnormal areas according to the temperature, pressure and flue gas flow data of the boiler heating surface, and select the overlapping parts of the abnormal areas to be marked as dust accumulation areas.

[0048] The specific analysis method of the abnormal area delineation module is as follows: the collected temperature data are sorted in sequence according to the positions of the measurement points, and the adjacent measurement points are grouped together. The temperature difference at each group of adjacent measurement points is calculated, and then the temperature difference is divided by the distance between each group of adjacent measurement points to obtain the temperature gradient between each group of adjacent measurement points. A temperature gradient standard value is set, and each group of adjacent measurement points with a temperature gradient greater than the temperature gradient standard value is screened out. According to their corresponding positions, the abnormal temperature change area is delineated; this helps to discover areas with uneven temperature distribution on the boiler heating surface.

[0049] The pressure difference at each group of adjacent measurement points is calculated, and the pressure gradient between each group of adjacent measurement points is obtained by dividing the pressure difference by the distance between each group of adjacent measurement points. A pressure gradient standard value is set, and each group of adjacent measurement points with a pressure gradient greater than the pressure gradient standard value is screened out. Based on their corresponding positions, an abnormal pressure change area is delineated; an abnormal pressure gradient may reflect problems such as uneven airflow distribution inside the boiler or pipe blockage. By delineating the abnormal pressure change area, these problems can be discovered and resolved in a timely manner to ensure the normal operation of the boiler.

[0050] Arrange the flow data in chronological order, calculate the flow change at each adjacent time point at each measuring point, and obtain the flow change rate at each adjacent time point at each measuring point by dividing it by the time interval. Subtract the flow change rate from the average flow change rate at each adjacent time point to obtain the flow change rate difference at each measuring point. Set a difference standard, filter out the measuring points with flow change rate differences greater than the difference standard, and delineate the flow abnormality area based on their corresponding positions. By delineating the flow abnormality area, we can understand the airflow conditions inside the boiler, optimize the combustion process, and improve the efficiency of the boiler.

[0051] The abnormal temperature change area, abnormal pressure change area, and abnormal flow area are read separately, and the overlapping area is selected and marked as the dust accumulation area; the position and range of the dust accumulation area can be accurately determined, providing an accurate target area for subsequent dust accumulation thickness and concentration analysis and soot blowing operations.

[0052] The dust accumulation area boundary determination module is used to search for edge points in the dust accumulation area to form a boundary contour of the dust accumulation area.

[0053] See also Figure 2 As shown, the specific analysis method of the dust accumulation area boundary determination module is: the first step is to select a point from the dust accumulation area as the starting point, starting from the starting point, find the adjacent points in the set direction, and obtain the temperature, pressure, and flow data of the starting point and the adjacent points respectively, and calculate the temperature change, pressure change, and flow change according to the obtained temperature, pressure, and flow data of the starting point and the adjacent points; when determining the edge point, the changes in the three parameters of temperature, pressure, and flow are comprehensively considered, avoiding the limitation of judging the edge point based on only a single parameter, and improving the accuracy and reliability of boundary determination.

[0054] In the second step, each change is compared with the corresponding preset threshold. If the change of any parameter between adjacent points exceeds the corresponding preset threshold, the adjacent point is identified as an edge point and the coordinates of the point are recorded. If the change of all parameters does not exceed the threshold, the adjacent point is judged not to be an edge point.

[0055] In the third step, after determining that the adjacent point is not an edge point, continue to search for the adjacent points of the adjacent point according to the above operation until a new edge point is found, and use the newly identified edge point as the current point, repeat the above operation until the search returns to the starting point; by starting from the starting point, continuously searching for adjacent points according to certain rules and judging whether they are edge points until the search returns to the starting point, the boundary of the dust accumulation area can be fully searched to ensure the integrity of the boundary contour.

[0056] The fourth step is to connect all edge points identified and recorded during the search process in a certain order to form the boundary outline of the dust accumulation area; this can accurately determine the boundary of the dust accumulation area and provide an accurate range for subsequent dust thickness and concentration analysis.

[0057] The thermal efficiency loss prediction module is used to calculate the ash accumulation thickness and concentration, build a thermal efficiency loss prediction model through simulation data, and calculate the thermal efficiency loss prediction value.

[0058] The specific analysis method for the ash thickness and ash concentration is as follows: a regular grid is constructed with the boundary contour of the ash accumulation area as the range, the grid intersection points are selected as each detection position, a coupling agent is applied to each selected detection position, an ultrasonic pulse is emitted by an ultrasonic transducer, the ultrasonic pulse is vertically incident on the ash layer, and the time from emission to reception of the reflected wave is recorded. The average value is obtained by multiple measurements to obtain the propagation time of the ultrasonic wave in the ash accumulation at each detection position, and the reflection intensity of the ultrasonic wave reflected back to the ultrasonic transducer at each detection position is obtained from the equipment. Combined with the propagation speed of the ultrasonic wave in the ash accumulation, the ash thickness at each detection position is accurately calculated, which can provide quantitative data on the ash thickness and provide a basis for evaluating the impact of ash accumulation on the heat transfer performance of the boiler.

[0059] The ash thickness at each detection location is calculated based on the propagation time and propagation speed of the ultrasonic wave in the ash at each detection location, and the ash concentration at each detection location is calculated based on the calculated ash thickness at each detection location and the reflection intensity of the ultrasonic wave reflected back to the ultrasonic transducer at each detection location. The calculation of the ash concentration helps to understand the nature and distribution of the ash accumulation and further analyze the impact of the ash accumulation on boiler operation.

[0060] It should be noted that the dust thickness is obtained by calculating the product of the propagation speed of the ultrasonic wave in the dust and half of the propagation time. The dust concentration is obtained by dividing the reflection intensity of the ultrasonic wave reflected back to the ultrasonic transducer at each detection position by The dust accumulation thickness at each detection position is obtained by multiplying is the constant affecting the reflection intensity by other factors.

[0061] Data on ash thickness and concentration are an important basis for subsequent prediction of thermal efficiency loss. By accurately measuring ash thickness and concentration, the impact of ash accumulation on boiler thermal efficiency can be more accurately assessed, providing data support for optimizing boiler operation and formulating reasonable soot blowing strategies.

[0062] The specific analysis method of the thermal efficiency loss prediction module is as follows: a series of ash thicknesses and ash concentrations are set at fixed intervals, and they are matched one-to-one to obtain various combinations of ash thicknesses and concentrations. For each combination of ash thicknesses and concentrations, a geometric model of the heat transfer system is constructed, and the grid is divided, and boundary conditions and initial conditions are set for it respectively. The heat flux density of the heat transfer surface is obtained according to Fourier's law, the heat transfer surface is integrated, and the heat transfer rate is calculated based on the geometric shape and heat flux density of the heat transfer surface. The module can simulate the heat transfer process under different ash deposition parameters, comprehensively understand the impact of ash deposition on heat transfer, and provide quantitative indicators for evaluating the impact of ash deposition on boiler thermal efficiency.

[0063] It should be noted that the calculation of heat transfer rate according to the geometric shape and heat flux density of the heat transfer surface is divided into planar heat transfer surface, cylindrical heat transfer surface and irregular geometric shape.

[0064] If the heat transfer surface is a plane, obtain the area and heat flux density of the heat transfer surface, and calculate the heat transfer rate by multiplying the area and heat flux density of the heat transfer surface.

[0065] If the heat transfer surface is a cylindrical surface, the radius and length of the cylindrical surface are obtained, and the lateral area of ​​the cylindrical surface is calculated. The heat transfer rate is obtained by calculating the product of the lateral area of ​​the cylindrical surface and the heat flux density.

[0066] If the heat transfer surface has an irregular geometric shape, divide the heat transfer surface into several grid units, calculate the product of the area of ​​each grid unit and the corresponding heat flux density, and sum them to obtain the heat transfer rate.

[0067] It should be noted that the boundary conditions are set as physical quantities on the model boundary, including inlet conditions: flue gas temperature, pressure, and flow velocity distribution.

[0068] Outlet conditions: pressure boundary (such as static pressure outlet), wall conditions: heat flux density (calculated by heat conduction of dust layer) and wall roughness (affecting flow boundary layer).

[0069] The initial conditions include full-field initialization: assigning initial temperature, pressure, and velocity values ​​to all nodes in the computational domain.

[0070] Iteration starting point: Usually the calculation results of a simplified model or empirical data are used as the initial field.

[0071] Accelerated convergence: Reasonable initial conditions can reduce the number of iterations and improve computational efficiency.

[0072] Under the condition of no dust accumulation, the heat transfer rate without dust accumulation is simulated and calculated, and the thermal efficiency loss under the current dust accumulation parameters is calculated according to the definition formula of thermal efficiency loss. In this way, each set of dust accumulation parameters is simulated and calculated to obtain the thermal efficiency loss under each combination of dust thickness and concentration.

[0073] It should be noted that the definition formula of the thermal efficiency loss is: ,in, represents the heat transfer rate under the condition of no dust accumulation, It represents the heat transfer rate under the current dust accumulation parameters (combination of dust accumulation thickness and concentration). The percentage of thermal efficiency loss is obtained by calculating the difference between the heat transfer rate in the absence of dust accumulation and the heat transfer rate in the current dust accumulation state and the ratio of the heat transfer rate in the absence of dust accumulation to the heat transfer rate in the absence of dust accumulation.

[0074] The heat transfer equation is established based on the first law of thermodynamics and the basic equations of fluid mechanics. The temperature field and velocity field distribution information are obtained by solving them, and then the heat flux density of the heat transfer surface is extracted according to Fourier's law. The Fourier law states that the heat flux density is proportional to the temperature gradient. The heat flux density is obtained by calculating the temperature gradient normal to the wall and multiplying it by the thermal conductivity of the dust accumulation.

[0075] The integration of the heat transfer surface is performed by multiplying the heat flux density of each small grid unit on the heat transfer surface by the area of ​​the unit, and then adding the results of all units. In this way, the heat transfer rate is calculated based on the geometric shape of the heat transfer surface and the heat flux density. The heat transfer rate when there is no dust accumulation is set as a reference value. By comparing the heat transfer rates when there is dust accumulation and when there is no dust accumulation, the thermal efficiency loss is calculated. The calculation is performed for each combination of dust accumulation thickness and concentration, and the relationship between the thermal efficiency loss and the dust accumulation parameters is established, thereby providing a basis for predicting the thermal efficiency loss.

[0076] A polynomial regression equation is constructed based on the thermal efficiency loss under various combinations of dust thickness and concentration obtained by simulation. The values ​​of various coefficients are calculated through the regression algorithm to determine the thermal efficiency loss prediction model, and the thermal efficiency loss prediction value is calculated accordingly.

[0077] It should be noted that the polynomial regression equation is about the dust accumulation thickness. and dust concentration The quadratic polynomial regression equation for : ,in are the coefficients to be determined respectively, and the partial derivatives of the coefficients to be determined are calculated by the least squares method, and the partial derivatives are set to 0 to obtain a set of linear equations about the coefficients to be determined. The values ​​of the coefficients are obtained by solving the set of equations, and the values ​​are substituted into the constructed quadratic polynomial regression equation to form a thermal efficiency loss prediction model.

[0078] The dust accumulation thickness and dust accumulation concentration at each detection position are substituted into the thermal efficiency loss prediction model to obtain the thermal efficiency loss prediction value at each detection position.

[0079] The dust accumulation core area determination module is used to compare the dust accumulation concentration, thickness and thermal efficiency loss with the set threshold value to screen out the dust accumulation core area.

[0080] See also Figure 3 As shown, the specific analysis method of the ash core area determination module is: the first step is to set thresholds for ash concentration, thickness and thermal efficiency loss respectively, read the ash thickness, ash concentration and thermal efficiency loss prediction value of each detection position, and compare them with the corresponding thresholds respectively. If the ash thickness, ash concentration and thermal efficiency loss prediction value of a certain detection position are greater than or equal to the corresponding threshold, the area where the detection position is located is preliminarily determined to be the ash core detection point; this helps to determine the area where ash accumulation has a greater impact on boiler performance, and provides a basis for focusing on and handling.

[0081] In the second step, with the dust core detection point as the center, the dust core detection point is judged for each vertex of its four adjacent grids respectively. If among the four vertices of a certain adjacent grid, there are dust thicknesses, dust concentrations, and thermal efficiency loss prediction values ​​greater than or equal to two vertices that are greater than or equal to the corresponding thresholds, then the grid is judged to belong to the dust core detection area, and the determined dust core detection areas are connected to obtain the dust core area; the scope of the dust core area can be accurately delineated, providing a target area for targeted soot blowing operations, thereby improving the soot blowing effect and efficiency.

[0082] The sootblowing control module is used to generate the sootblowing path in the soot accumulation core area, determine whether there is a sootblowing problem based on the temperature data before and after sootblowing, and calculate the angle at which the sootblower blowing head needs to be adjusted.

[0083] The specific analysis method of the soot blowing path in the soot accumulation core area is as follows: obtain the position and range of the soot accumulation core area, divide it into several sub-areas according to the shape of the soot accumulation core area, determine the blowing order of each sub-area according to the set logic, plan the soot blowing path according to the shape of each sub-area, connect the soot blowing paths planned in each sub-area to form a continuous soot blowing path, thereby controlling the soot blower to perform soot blowing; determine the blowing order of each sub-area according to the set logic, ensure the orderly progress of the soot blowing operation, improve the soot blowing efficiency, avoid repeated blowing or missing blowing areas, and form a continuous soot blowing path by connecting the soot blowing paths planned in each sub-area. The soot accumulation core area can be effectively covered, the comprehensiveness and accuracy of soot blowing can be improved, and the soot accumulation can be effectively removed.

[0084] It should be noted that the specific analysis method for planning the sootblowing path according to the shape of each sub-area is as follows: for rectangular sub-areas, a round-trip path is adopted, starting from one side of the rectangle, blowing in a direction parallel to the side, and returning after reaching the opposite side. The distance of each movement is determined according to the coverage width of the sootblower. At the same time, there needs to be a certain overlap between adjacent blowing paths, and the overlapping width is 10%-20% of the coverage width to avoid missing dust accumulation.

[0085] For circular sub-areas, a spiral or annular path can be selected. The spiral path starts from near the center of the circle and gradually expands outward along the spiral line. The annular path blows around the circumference of the circular sub-area, moving inward or outward in circles. The moving distance is determined by the coverage range of the soot blower, which is half of the coverage radius.

[0086] For triangular sub-areas, you can start from one vertex of the triangle, blow in a direction parallel to the base, and then gradually move upward until the entire triangular area is covered, or blow radially from the center of the triangle to the three sides. The angle of each blow and the distance of movement are determined according to the coverage angle of the soot blower and the length of the triangle.

[0087] The specific analysis method for determining whether there is a soot blowing problem is as follows: several soot blowing detection points are selected in the core area of ​​boiler soot accumulation, and the temperature of each soot blowing detection point before and after soot blowing is respectively obtained by a temperature sensor. The temperature change rate of each soot blowing detection point is obtained by taking the difference and then dividing it by the soot blowing time. If the temperature change rate of a soot blowing detection point is positive and reaches a set threshold, it is determined that the soot accumulation is reduced and the heat transfer effect is improved. If the temperature change rate of a soot blowing detection point does not reach the set threshold, it is determined that there is a soot blowing problem and the soot blowing angle needs to be adjusted. If the temperature change rate of a soot blowing detection point does not reach the set threshold, it means that the soot accumulation at the detection point has not been effectively cleared, there is a soot blowing problem, and parameters such as the soot blowing angle need to be adjusted. This method can timely discover problems in the soot blowing process, provide a basis for optimizing the soot blowing operation, and improve the effect and efficiency of soot blowing.

[0088] The specific analysis method for the angle that the sootblower blowing head needs to be adjusted is: obtaining the sub-area to which the sootblowing detection point with the sootblowing problem belongs, obtaining the width of the soot accumulation area of ​​the sub-area, measuring the width of the soot accumulation area that can be covered by the current sootblowing angle, subtracting to obtain the width of the soot accumulation area not covered by the current sootblowing angle, and combining the sootblower blowing coverage width to calculate the angle that the sootblower blowing head needs to be adjusted, so as to adjust the sootblower blowing head; by accurately calculating the adjustment angle, the sootblower can more effectively cover the soot accumulation area, improve the sootblowing effect, and reduce the impact of soot accumulation on boiler operation.

[0089] It should be noted that the specific analysis method for the angle that needs to be adjusted for the sootblower head is as follows: the sootblower sweep coverage width and the width of the dust accumulation area not covered by the current sootblowing angle are respectively recorded as , through the formula Calculate the angle that the sootblower head needs to be adjusted .

[0090] Management database, used to store measurement data, abnormal area data, dust accumulation area data, thermal efficiency loss prediction model data and soot blowing control data.

[0091] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention, which are still covered by the scope of protection of the present invention.

Claims

1. A boiler automatic sootblowing system based on three-dimensional finite element neural network, characterized in that: The system specifically includes the following modules: Data acquisition module, used to obtain the temperature, pressure and flue gas flow data of the boiler heating surface respectively; The abnormal area delineation module is used to screen abnormal areas based on the temperature, pressure and flue gas flow data of the boiler heating surface, and select the overlapping parts of the abnormal areas and mark them as dust accumulation areas; A dust accumulation area boundary determination module is used to search for edge points in the dust accumulation area to form a boundary contour of the dust accumulation area; Thermal efficiency loss prediction module, used to calculate the thickness and concentration of ash accumulation, build a thermal efficiency loss prediction model through simulation data, and calculate the predicted value of thermal efficiency loss; The dust accumulation core area determination module is used to compare the dust accumulation concentration, thickness and thermal efficiency loss with the set threshold value to screen out the dust accumulation core area; The sootblowing control module is used to generate the sootblowing path in the core area of ​​soot accumulation, determine whether there is a sootblowing problem based on the temperature data before and after sootblowing, and calculate the angle at which the sootblower blowing head needs to be adjusted; Management database, used to store measurement data, abnormal area data, dust accumulation area data, thermal efficiency loss prediction model data and soot blowing control data.

2. The boiler automatic sootblowing system based on three-dimensional finite element neural network according to claim 1 is characterized in that: The specific analysis method of the data acquisition module is: Several equally spaced measurement points are selected on the surface of the boiler heating surface. The temperature and pressure of each measurement point are obtained by temperature sensors and pressure sensors respectively. At the same time, the time points are divided into equal intervals, and the flue gas flow rate at each measurement point at each time point is detected to obtain the flow rate at each measurement point at each time point.

3. The boiler automatic sootblowing system based on three-dimensional finite element neural network according to claim 2, characterized in that: The specific analysis method of the abnormal area delineation module is: The collected temperature data are sorted in order by the position of the measurement points. Adjacent measurement points are grouped together. The temperature difference at each group of adjacent measurement points is calculated, and then the temperature difference is divided by the distance between each group of adjacent measurement points to obtain the temperature gradient between each group of adjacent measurement points. A temperature gradient standard value is set, and each group of adjacent measurement points with a temperature gradient greater than the temperature gradient standard value is screened out. Based on their corresponding positions, the temperature change abnormal area is delineated; Calculate the pressure difference between each group of adjacent measurement points, divide the pressure difference by the distance between each group of adjacent measurement points to obtain the pressure gradient between each group of adjacent measurement points, set a pressure gradient standard value, screen out each group of adjacent measurement points whose pressure gradient is greater than the pressure gradient standard value, and delineate the abnormal pressure change area based on their corresponding positions; Arrange the flow data in chronological order, calculate the flow change at each adjacent time point at each measuring point, divide it by the time interval to obtain the flow change rate at each adjacent time point at each measuring point, and subtract it from the average flow change rate at each adjacent time point to obtain the flow change rate difference at each measuring point. Set a difference standard, filter out the measuring points with a flow change rate difference greater than the difference standard, and delineate the flow abnormality area based on their corresponding location; Read the abnormal temperature change area, abnormal pressure change area, and abnormal flow area respectively, and select the overlapping area to mark as the dust accumulation area.

4. The boiler automatic sootblowing system based on three-dimensional finite element neural network according to claim 1, characterized in that: The specific analysis method of the dust accumulation area boundary determination module is as follows: The first step is to select a point in the dust accumulation area as the starting point, start from the starting point, search for adjacent points in the set direction, and obtain the temperature, pressure, and flow data of the starting point and the adjacent points respectively. Based on the obtained temperature, pressure, and flow data of the starting point and the adjacent points, the temperature change, pressure change, and flow change are calculated; In the second step, each change is compared with the corresponding preset threshold. If the change of any parameter between adjacent points exceeds the corresponding preset threshold, the adjacent point is identified as an edge point and the coordinates of the point are recorded. If the change of all parameters does not exceed the threshold, the adjacent point is determined not to be an edge point. In the third step, after determining that the adjacent point is not an edge point, continue to search for the adjacent point of the adjacent point according to the above operation until a new edge point is found, and use the newly identified edge point as the current point, repeating the above operation until the search returns to the starting point; The fourth step is to connect all edge points identified and recorded during the search process in a certain order to form the boundary outline of the dust accumulation area.

5. The boiler automatic sootblowing system based on three-dimensional finite element neural network according to claim 4 is characterized in that: The specific analysis method of the dust accumulation thickness and dust accumulation concentration is: A regular grid is constructed based on the boundary outline of the dust accumulation area. The intersection points of the grid are selected as the test locations. Coupling agent is applied to each selected test location. An ultrasonic transducer transmits an ultrasonic pulse perpendicularly into the dust layer. The duration from emission to reception of the reflected wave is recorded. The average value is calculated through multiple measurements to determine the propagation time of the ultrasonic wave in the dust accumulation at each test location. The reflection intensity of the ultrasonic wave reflected back to the ultrasonic transducer at each test location is also obtained from the device. The dust thickness at each detection position is calculated based on the propagation time and propagation speed of the ultrasonic wave in the dust at each detection position, and the dust concentration at each detection position is calculated based on the calculated dust thickness at each detection position and the reflection intensity of the ultrasonic wave reflected back to the ultrasonic transducer at each detection position.

6. The boiler automatic sootblowing system based on three-dimensional finite element neural network according to claim 1, characterized in that: The specific analysis method of the thermal efficiency loss prediction module is: A series of dust accumulation thicknesses and concentrations are set at fixed intervals and mapped one-to-one to obtain various combinations of dust accumulation thickness and concentration. For each combination of dust accumulation thickness and concentration, a geometric model of the heat transfer system is constructed and meshed, with boundary conditions and initial conditions set for each. The heat flux density of the heat transfer surface is obtained based on Fourier's law, and the heat transfer surface is integrated. Based on the geometry of the heat transfer surface and the heat flux density, the heat transfer rate is calculated. Under the condition of no dust accumulation, the heat transfer rate without dust accumulation is simulated and calculated, and the thermal efficiency loss under the current dust accumulation parameters is calculated according to the definition formula of thermal efficiency loss. In this way, the thermal efficiency loss under each set of dust accumulation parameters is simulated and calculated to obtain the thermal efficiency loss under each combination of dust accumulation thickness and concentration; A polynomial regression equation is constructed based on the thermal efficiency loss under various combinations of dust thickness and concentration obtained by simulation. The values ​​of each coefficient are calculated using a regression algorithm to determine the thermal efficiency loss prediction model, and the thermal efficiency loss prediction value is calculated based on this. The dust accumulation thickness and dust accumulation concentration at each detection position are substituted into the thermal efficiency loss prediction model to obtain the thermal efficiency loss prediction value at each detection position.

7. The boiler automatic sootblowing system based on three-dimensional finite element neural network according to claim 1, characterized in that: The specific analysis method of the dust accumulation core area determination module is: The first step is to set thresholds for ash concentration, thickness, and thermal efficiency loss, respectively. The ash thickness, concentration, and predicted thermal efficiency loss values ​​at each detection location are read and compared with the corresponding thresholds. If the ash thickness, concentration, and predicted thermal efficiency loss values ​​at a detection location are all greater than or equal to the corresponding thresholds, the area where the detection location is located is preliminarily determined to be a core ash detection point. In the second step, with the dust core detection point as the center, the dust core detection point is judged for each vertex of its four adjacent grids. If among the four vertices of a certain adjacent grid, there are dust thicknesses, dust concentrations, and thermal efficiency loss prediction values ​​greater than or equal to two vertices that are greater than or equal to the corresponding thresholds, then the grid is judged to belong to the dust core detection area. In this way, the determined dust core detection areas are connected to obtain the dust core area.

8. The boiler automatic sootblowing system based on three-dimensional finite element neural network according to claim 1, characterized in that: The specific analysis method of the soot blowing path in the soot accumulation core area is as follows: Obtain the location and range of the dust accumulation core area, divide it into several sub-areas according to its shape, determine the blowing order of each sub-area according to the set logic, plan the soot blowing path according to the shape of each sub-area, and connect the soot blowing paths planned in each sub-area to form a continuous soot blowing path, so as to control the soot blower to perform soot blowing.

9. The boiler automatic sootblowing system based on three-dimensional finite element neural network according to claim 8, characterized in that: The specific analysis method for determining whether there is a sootblowing problem is as follows: Several soot blowing detection points are selected in the core area of ​​boiler soot accumulation. The temperature of each soot blowing detection point before and after soot blowing is obtained through a temperature sensor. The temperature change rate of each soot blowing detection point is obtained by taking the difference and dividing it by the soot blowing time. If the temperature change rate of a soot blowing detection point is positive and reaches the set threshold, it is judged that the soot accumulation is reduced and the heat transfer effect is improved. If the temperature change rate of a soot blowing detection point does not reach the set threshold, it is judged that there is a soot blowing problem and the soot blowing angle needs to be adjusted.

10. The boiler automatic sootblowing system based on three-dimensional finite element neural network according to claim 9, characterized in that: The specific analysis method of the angle that the sootblower blowing head needs to adjust is: Obtain the sub-area to which the sootblowing detection point with the sootblowing problem belongs, obtain the width of the soot accumulation area in the sub-area, measure the width of the soot accumulation area that can be covered by the current sootblowing angle, and obtain the width of the soot accumulation area not covered by the current sootblowing angle by subtraction. Combined with the sootblower blowing coverage width, calculate the angle that the sootblower blowing head needs to be adjusted, and adjust the sootblower blowing head accordingly.

Images