Online calculation method and device for air volume distribution of boiler secondary air bellow and computer equipment

Abstract

The invention discloses an online calculation method and device for air distribution of a boiler secondary air bellow and computer equipment, belongs to the technical field of numerical simulation and machine learning, and particularly relates to online calculation and monitoring of air distribution of the boiler secondary air bellow. The problems that in the prior art, a sensor measurement defect (for example, a measurement error caused by airflow turbulence in a secondary air duct cannot be overcome) and a big data sub-monitoring algorithm defect (for example, once sensor data is abnormal, a big data analysis module has calculation deviation due to lack of mechanical constraint) are solved; the method comprises the steps that a trained online computing machine learning model is arranged in a boiler DCS control system, and by inputting real-time operation data, the air volume of a burner is calculated online and output to an operation picture of the boiler DCS control system for an operator to use. The on-line calculation method and device for the air volume distribution of the boiler secondary air bellow and the computer equipment are suitable for on-line calculation and monitoring of the air volume distribution of the boiler secondary air bellow.

Description

Technical Field

[0001] The present invention relates to the technical field of numerical simulation and machine learning, and in particular to online calculation and monitoring of air volume distribution of a boiler secondary wind box. Background Art

[0002] Boiler secondary air distribution is crucial to combustion efficiency, pollutant control and safe production.

[0003] At present, the measurement of secondary air box air volume distribution mainly involves installing sensors for monitoring. However, the following problems exist in engineering applications:

[0004] 1) The sensor requires a long straight pipe section at the installation location, but the straight pipe section of the secondary air duct is usually short. Generally, the requirement can be met on the main air duct, but it can only partially meet the requirements for each burner nozzle, and it cannot accurately measure the air volume distribution of each burner nozzle;

[0005] 2) The secondary air contains dust. After the sensor has been running for a period of time, the dust will wear and clog the sensor's pressure-inducing holes and other small structures, causing it to malfunction.

[0006] 3) The airflow in the secondary air duct is turbulent, and the flow velocity at the same section varies greatly. The sensor can only measure the local flow velocity, and the measurement results fluctuate greatly and cannot reflect the overall flow characteristics.

[0007] Due to the existence of the above problems, the sensor-based secondary air box air volume distribution measurement results are inaccurate. When the boiler has problems such as overheating, high pollutant emissions, coking, wear, etc., it cannot provide an effective reference for the precise control of boiler operation.

[0008] At present, in order to overcome the above problems, technical personnel in this field have carried out some research work on the air volume monitoring technology of the secondary air box of coal-fired power plant boilers. The results can be mainly divided into two categories:

[0009] (1) Develop new wind volume sensors, such as using sound waves to measure wind volume and wind speed in real time.

[0010] (2) Through big data analysis algorithms, monitor the secondary air volume distribution from the perspective of secondary air control and air volume prediction.

[0011] The application of new sensors has improved measurement sensitivity and interference resistance to a certain extent, but it cannot overcome the measurement errors caused by the turbulent airflow in the secondary air duct, and its improvement effect on traditional Venturi flow measurement devices is limited. This is mainly due to the following reasons: the secondary air duct is compact and contains a large amount of turbulent airflow. The flow velocity varies greatly within the same cross-section, and the measurement equipment can only measure the wind speed at a certain location, unable to obtain the average velocity of the cross-section. In addition, the presence of turbulent gas causes the sensor's measurement value to jump, affecting the acquisition of a reference value.

[0012] The research results of realizing the prediction and control of secondary air volume through big data analysis algorithms are mainly based on the analysis and mining of sensor measurement data, which has improved the adaptive adjustment of the secondary air control system to a certain extent. However, once the sensor data is abnormal, the big data analysis module will have calculation deviations due to the lack of mechanistic constraints, and its reliability needs to be improved.

[0013] In summary, there is an urgent need for a secondary air online monitoring method that can overcome the shortcomings of existing sensor measurements and the defects of big data monitoring algorithms. Summary of the Invention

[0014] The present invention proposes an online calculation method, device and computer equipment for the air volume distribution of the boiler secondary wind box, which solves the problems of sensor measurement drawbacks in the prior art (such as the inability to overcome measurement errors caused by airflow turbulence in the secondary air duct) and defects in the big data analysis monitoring algorithm (such as once the sensor data is abnormal, the big data analysis module will have calculation deviations due to the lack of mechanistic constraints).

[0015] The online calculation method for boiler secondary air box air volume distribution of the present invention comprises:

[0016] Step S1: Collect boiler operation data and structural data; form an operation data set DO based on the operation data; construct a three-dimensional geometric model of the secondary wind box and burner based on the structural data;

[0017] Step S2: Based on the operating data set DO and the burner's three-dimensional geometric model, numerical simulation calculations are performed on the pressure drop of a single burner under different air volume conditions to obtain a pressure drop data set PD, where the pressure drop statistical boundaries are the burner inlet and outlet. The pressure drop data set PD includes the pressure drops of n typical boiler operating conditions.

[0018] Step S3: Simplify the burner structure in the three-dimensional geometric model of the secondary wind box, establish a mapping relationship between the pressure drop in the pressure drop dataset PD and the simplified burner structure, and calculate the numerical correspondence between the pressure drop and the viscous drag coefficient and inertial drag coefficient in the porous medium region under the same burner mass flow inlet boundary conditions to form a drag coefficient dataset PR;

[0019] Step S4: Based on the operating data set DO and the drag coefficient data set PR, the least squares method is used to fit the functional relationship between the air volume and the viscous drag coefficient and the inertial drag coefficient respectively by minimizing the sum of squares of the errors, thereby obtaining the fitted drag coefficient data set FC;

[0020] Step S5: The simplified burner structure area in the three-dimensional geometric model of the secondary air box after the burner structure is simplified is used as the resistance area. The damper in the air box is set to be open. The data in the running data set DO is used as the boundary condition to perform the secondary air volume distribution calculation to obtain the secondary air box air volume distribution calculation result data set DS.

[0021] Step S6: using an extreme learning machine to train the numerical mapping relationship between the operating data in the operating data set DO and the secondary air box air volume distribution calculation result data set DS, to obtain a trained online computing machine learning model;

[0022] Step S7: The trained online machine learning model is placed in the boiler DCS control system. By inputting real-time operating data, the burner air volume is calculated online and output to the operation screen of the boiler DCS control system for use by the operator.

[0023] Furthermore, a preferred embodiment is provided, wherein the operating data includes secondary air volume SA, secondary air temperature ST, damper opening state SD, and air volume measurement data SM;

[0024] Run the dataset DO:

[0025]

[0026] Among them, [SA i ST i SD i SM i ] represents the operating measurement point data of the typical operating condition i of the boiler, and n is the total number of typical operating conditions of the boiler.

[0027] Furthermore, a preferred embodiment is provided, wherein step S2 comprises:

[0028] Finite element analysis is used to mesh the three-dimensional geometric model of the burner and verify its mesh independence.

[0029] In the burner 3D geometric model after meshing, the operating point data of the typical boiler operating condition i [SA i ST i SD i SM i] as the boundary condition, the numerical simulation calculation of the pressure drop of a single burner under different air volume conditions is carried out based on the mass conservation, momentum conservation and energy conservation equations;

[0030] The pressure drop statistical boundaries are the burner inlet and outlet, forming the pressure drop data set PD.

[0031] Furthermore, a preferred embodiment is provided, wherein step S4 includes:

[0032] Considering the nonlinear relationship between the burner air volume and the viscous drag coefficient and the inertial drag coefficient, the least squares method is used to fit a quadratic polynomial; the form of the quadratic polynomial is:

[0033] y=a+bx+cx 2

[0034] Where a, b, and c are the coefficients of the constant term, the linear term, and the quadratic term, respectively; x is the mass flow rate, and y is the pressure drop of the burner; y comes from the resistance coefficient dataset PR, and x comes from the operating dataset DO;

[0035] The sum of squared errors is defined as:

[0036]

[0037] By taking the derivative and setting it to 0, we can find the values ​​of a, b, and c:

[0038]

[0039] Finally, the fitted drag coefficient data set FC is formed.

[0040] Furthermore, a preferred embodiment is provided, wherein the secondary air volume distribution calculation includes:

[0041] Statistical inlet mass flow rate before the resistance area (i.e. the area where the simplified burner structure is located);

[0042] The resistance coefficient is set by fitting the resistance coefficient data set FC, and the air volume distribution of the secondary air box is numerically simulated.

[0043] The mass flow rates at all burner nozzle outlets are counted to ultimately form a secondary wind box air volume distribution calculation result data set DS.

[0044] Furthermore, a preferred embodiment is provided, wherein step S6 includes:

[0045] Step 6.1: The secondary air box air volume distribution calculation result dataset DS is combined with the operating data in the operating dataset DO (i.e., secondary air volume SA, secondary air temperature ST, and damper opening state SD) to form basic modeling data, where the secondary air box air volume distribution calculation result dataset DS is the target value;

[0046] Step 6.2: For the basic modeling data, considering the dimensional differences of different parameters, perform max-min normalization on the data:

[0047]

[0048] Among them, x i is the sample value of any parameter in the basic modeling data, x * is the normalized value, x min 、x max are the minimum and maximum values ​​of any parameter sample point respectively;

[0049] Step 6.3: Input the normalized basic modeling data into the ELM neural network calculation model, improve the accuracy of the model calculation by adjusting the parameters, and evaluate the model after denormalizing the calculation results. Based on the evaluation results, select the model with the highest accuracy as the trained online computing machine learning model.

[0050] The present invention also proposes an online calculation device for boiler secondary wind box air volume distribution, the device comprising:

[0051] Module S1: Collect boiler operation data and structural data; generate an operation data set DO based on the operation data; construct a three-dimensional geometric model of the secondary air box and burner based on the structural data;

[0052] Module S2: Based on the operating data set DO and the burner's three-dimensional geometric model, numerical simulation calculations are performed on the pressure drop of a single burner under different air volume conditions to obtain a pressure drop data set PD. The pressure drop statistical boundaries are the burner inlet and outlet. The pressure drop data set PD includes the pressure drops of n typical boiler operating conditions.

[0053] Module S3: Simplify the burner structure in the 3D geometric model of the secondary air box, establish a mapping relationship between the pressure drop in the pressure drop dataset PD and the simplified burner structure, and calculate the numerical correspondence between the pressure drop and the viscous drag coefficient and inertial drag coefficient in the porous media region under the same burner mass flow inlet boundary conditions to form the drag coefficient dataset PR.

[0054] Module S4: Based on the operating data set DO and the drag coefficient data set PR, the least squares method is used to fit the functional relationship between the air volume and the viscous drag coefficient and the inertial drag coefficient by minimizing the sum of squared errors, and the fitted drag coefficient data set FC is obtained;

[0055] Module S5: The simplified burner structure area in the simplified burner structure 3D geometric model of the secondary air box is used as the resistance area. The damper in the air box is set to be open. The data in the running data set DO is used as the boundary condition to perform the secondary air volume distribution calculation and obtain the secondary air box air volume distribution calculation result data set DS.

[0056] Module S6: Using an extreme learning machine to train the numerical mapping relationship between the operating data in the operating data set DO and the secondary air box air volume distribution calculation result data set DS, to obtain a trained online computing machine learning model;

[0057] Module S7: Place the trained online machine learning model in the boiler DCS control system. By inputting real-time operating data, calculate the burner air volume online and output it to the operation screen of the boiler DCS control system for use by operators.

[0058] The present invention also proposes a computer device, comprising: a processor and a memory, wherein the memory is used to store executable instructions of the processor, and the processor is configured to execute any one of the above-mentioned online calculation methods for boiler secondary wind box air volume distribution by executing the executable instructions.

[0059] The present invention also proposes a computer storage medium, in which a computer program is stored. When the computer program is run, any one of the above-mentioned online calculation methods for boiler secondary wind box air volume distribution is executed.

[0060] The present invention also proposes a computer program product, comprising a computer program / instruction, which, when executed by a processor, implements the steps of any one of the above-mentioned methods for online calculation of boiler secondary wind box air volume distribution.

[0061] The present invention has the following beneficial effects:

[0062] 1. The online calculation method for the air volume distribution of the boiler secondary wind box described in the present invention is based on a method that combines numerical simulation and machine learning to realize the online calculation of the air volume distribution of the boiler secondary wind box. Based on the measurement data of the existing sensors of the boiler, numerical simulation calculation is used to establish the mechanism gap between the damper and the air volume distribution. The big data analysis method is used to perform data mining on the damper and air volume distribution to realize the online calculation and monitoring of the secondary air volume.

[0063] 2. The online calculation method for the boiler secondary air box air volume distribution described in the present invention adopts a method that combines numerical simulation and machine learning to realize online monitoring of the secondary air volume distribution, providing an online monitoring means for operating personnel to timely grasp the boiler combustion status, avoiding problems such as overheating, high pollutant emissions, coking, wear, etc., and providing an effective reference for the precise control of boiler operation. It solves the problems of sensor measurement drawbacks in the existing technology (such as the inability to overcome measurement errors caused by airflow turbulence in the secondary air duct) and defects in big data monitoring algorithms (such as once the sensor data is abnormal, the big data analysis module will have calculation deviations due to the lack of mechanistic constraints).

[0064] 3. The online calculation method for boiler secondary air box air volume distribution described in the present invention helps operators to timely grasp the boiler combustion status through online monitoring of secondary air, provide reasonable optimization strategies based on the actual operating status, and provide operating adjustment suggestions to operators, thereby improving the safety and economy of unit operation and promoting safe and efficient production of power plants.

[0065] The method, device and computer equipment for online calculation of boiler secondary wind box air volume distribution of the present invention are suitable for online calculation and monitoring of boiler secondary wind box air volume distribution. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in 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 paying any creative work.

[0067] Figure 1 The figure is a flow chart of an online calculation method for boiler secondary wind box air volume distribution in one embodiment of the present invention. DETAILED DESCRIPTION

[0068] In order to make the technical solutions and advantages of the present invention more clearly described, the specific embodiments of the present invention will be further described in detail and completely in conjunction with the accompanying drawings. The various embodiments described below are only part of the preferred embodiments of the present invention, rather than all implementation plans; the various embodiments described below are intended to explain the present invention and cannot be understood as limiting the present invention; the reasonable combination of the technical features defined in the various embodiments of the present invention, as well as all other implementation plans obtained by ordinary technicians in this field based on the embodiments of the present invention without making creative work, all fall within the scope of protection of the present invention.

[0069] Implementation 1: An online calculation method for boiler secondary air box air volume distribution is provided, the method comprising the following steps:

[0070] Step S1: Collect boiler operation data and structural data; form an operation data set DO based on the operation data; construct a three-dimensional geometric model of the secondary wind box and burner based on the structural data;

[0071] Step S2: Based on the operating data set DO and the burner's three-dimensional geometric model, numerical simulation calculations are performed on the pressure drop of a single burner under different air volume conditions to obtain a pressure drop data set PD, where the pressure drop statistical boundaries are the burner inlet and outlet. The pressure drop data set PD includes the pressure drops of n typical boiler operating conditions.

[0072] Step S3: Simplify the burner structure in the three-dimensional geometric model of the secondary wind box, establish a mapping relationship between the pressure drop in the pressure drop dataset PD and the simplified burner structure, and calculate the numerical correspondence between the pressure drop and the viscous drag coefficient and inertial drag coefficient in the porous medium region under the same burner mass flow inlet boundary conditions to form a drag coefficient dataset PR;

[0073] Step S4: Based on the operating data set DO and the drag coefficient data set PR, the least squares method is used to fit the functional relationship between the air volume and the viscous drag coefficient and the inertial drag coefficient respectively by minimizing the sum of squares of the errors, thereby obtaining the fitted drag coefficient data set FC;

[0074] Step S5: The simplified burner structure area in the three-dimensional geometric model of the secondary air box after the burner structure is simplified is used as the resistance area. The damper in the air box is set to be open. The data in the running data set DO is used as the boundary condition to perform the secondary air volume distribution calculation to obtain the secondary air box air volume distribution calculation result data set DS.

[0075] Step S6: using an extreme learning machine to train the numerical mapping relationship between the operating data in the operating data set DO and the secondary air box air volume distribution calculation result data set DS, to obtain a trained online computing machine learning model;

[0076] Step S7: The trained online machine learning model is placed in the boiler DCS control system. By inputting real-time operating data, the burner air volume is calculated online and output to the operation screen of the boiler DCS control system for use by the operator.

[0077] In this embodiment, the operation data is pre-processed to form an operation data set DO.

[0078] The preprocessing includes screening and cleaning the running data to remove abnormal values ​​and missing values.

[0079] In this embodiment, the operating data is used to formulate boundary conditions for the simulation calculation of the secondary air box air volume distribution.

[0080] In this embodiment, the three-dimensional geometric models of the secondary air box and the burner are further used for numerical simulation calculations.

[0081] In this embodiment, considering the nonlinear relationship between the burner air volume and the viscous drag coefficient and the inertial drag coefficient, the least squares method is used to fit the functional relationship between the air volume and the viscous drag coefficient and the inertial drag coefficient respectively by minimizing the sum of squares of errors.

[0082] In this embodiment, in the secondary air volume distribution calculation, secondary air volume distribution calculations for n typical working conditions are performed.

[0083] In this embodiment, the extreme learning machine is called “Extreme Learning Machine” in English.

[0084] In this embodiment, the numerical mapping relationship between the operating data in the training operating data set DO (including the secondary air volume SA, the secondary air temperature ST, and the damper opening state SD) and the secondary air box air volume distribution calculation result data set DS is:

[0085] The numerical mapping relationship between the operating data in the training operation data set DO (including secondary air volume SA, secondary air temperature ST, and air damper opening state SD) and the air volume distribution of each burner.

[0086] In this embodiment, the operating data in the operating data set DO includes: secondary air volume SA, secondary air temperature ST, and damper opening state SD.

[0087] In this embodiment, the real-time operation data includes: real-time secondary air volume SA, secondary air temperature ST, and air door damper opening state SD.

[0088] Implementation method 2: The operation data includes the secondary air volume SA, the secondary air temperature ST, the damper opening state SD, and the air volume measurement data SM;

[0089] Run the dataset DO:

[0090]

[0091] Among them, [SA i ST i SD i SM i ] represents the operating measurement point data of the typical operating condition i of the boiler, and n is the total number of typical operating conditions of the boiler.

[0092] In this embodiment:

[0093] SA i Characterize the secondary air volume of the boiler under typical operating condition i;

[0094] ST i Characterizes the secondary air temperature of the boiler under typical operating condition i;

[0095] SD i Characterizes the air damper opening state of the typical boiler operating condition i;

[0096] SM i Air volume measurement data representing a typical boiler operating condition i.

[0097] The secondary air volume SA, secondary air temperature ST, damper opening state SD and air volume measurement data SM are data obtained after preprocessing (such as screening and cleaning) of the original data measured at a given sensor measuring point, that is, operating data.

[0098] In this embodiment, the secondary air volume SA is the sum of the values ​​measured by the air volume sensors at the left and right outlets of the air preheater.

[0099] In this embodiment, the secondary air temperature ST is the average value measured by the air temperature sensor at the outlet of the air preheater.

[0100] In this embodiment, the damper opening state SD is a data set, SD = [OP1...OP i ...OP m ], where: OP i is the opening value of the i-th damper, with a total of m dampers, and the dampers correspond one to one with the burner nozzles.

[0101] In this embodiment, the secondary wind box air volume measurement sensor is generally arranged at the air duct and burner nozzle position. Since the gas turbulence phenomenon at the burner nozzle position is relatively serious, the sensor measurement point data located on the straight pipe section is selected, and it is screened and cleaned. After eliminating invalid data, the air volume measurement data SM is formed.

[0102] Implementation method 3: Step S2 includes:

[0103] Finite element analysis is used to mesh the three-dimensional geometric model of the burner and verify its mesh independence.

[0104] In the burner 3D geometric model after meshing, the operating point data of the typical boiler operating condition i [SA i ST i SD i SM i ] as the boundary condition, the numerical simulation calculation of the pressure drop of a single burner under different air volume conditions is carried out based on the mass conservation, momentum conservation and energy conservation equations;

[0105] The pressure drop statistical boundaries are the burner inlet and outlet, forming the pressure drop data set PD.

[0106] In this embodiment, the data dimension of the pressure drop data set PD is b×n, where n is the number of typical operating conditions and b is the number of types of burners.

[0107] Implementation 4: Step S3 specifically:

[0108] Finite element analysis was used to mesh the burner nozzle area in the three-dimensional geometric model of the secondary wind box, and its mesh independence was verified.

[0109] The burner structure is simplified in the meshed three-dimensional geometric model of the secondary air box, and a mapping relationship between the pressure drop in the pressure drop dataset PD and the simplified burner structure is established. The numerical correspondence between the pressure drop and the viscous drag coefficient and inertial drag coefficient in the porous medium region under the same burner mass flow inlet boundary conditions is calculated to form the resistance coefficient dataset PR.

[0110] In this embodiment, the data dimension of the resistance coefficient dataset PR is b×n, where n is the number of typical operating conditions and b is the number of types of burners.

[0111] In this embodiment, a custom function (or first custom function) in the numerical simulation software is used to calculate the numerical correspondence between the pressure drop and the viscous resistance coefficient and the inertial resistance coefficient in the porous medium area under the same burner mass flow inlet boundary conditions to form a resistance coefficient data set PR.

[0112] The user-defined function (or first user-defined function):

[0113] Through the user-defined function (UDF) interface provided by the commercial software ANSYS, a DEFINE_FROFILE type function was customized; the parameters involved are: chage_mass, thread, i, and the meanings of the parameters are: function name, surface pointer, and number of assignments respectively; the calculation function (the first user-defined function) is expressed as: mass = DEFINE_FROFILE (thread, i). Through the DEFINE_FROFILE type function named chage_mass, the inlet mass flow rate is simulated and calculated at different calculation time points, and then submitted to the commercial finite element analysis software ANSYS for simulation solution.

[0114] In this embodiment, the simplified burner structure:

[0115] All geometric features of the original burner were removed, and the numerical simulation calculation area was reconstructed to match the burner's outer contour dimensions. Finite element discretization was performed using a structured mesh, ignoring the impact of the burner's complex structure. A porous medium was used in the simulation to match the pressure drop caused by the burner's actual structure. This simplification significantly reduced the number of finite elements and improved the mesh quality, reducing computational time and improving accuracy.

[0116] In this embodiment, in step S3, a mapping relationship between the pressure drop in the pressure drop dataset PD and the simplified burner structure is established, that is, the pressure drop caused by the porous medium and the actual burner structure is matched by an iterative calculation method:

[0117] In porous media, under any given flow boundary condition, by adjusting the viscous drag coefficient and the inertial drag coefficient, the accuracy of the error (set as a) between the pressure drop PDK of the porous media in the simplified burner structure and the pressure drop PSJ caused by the actual burner structure meets the accuracy requirements;

[0118] Among them, the accuracy of the error is above 99.9% and is considered to meet the accuracy requirements;

[0119] The calculation method of the error accuracy is: a=|(PSJ-PDK) / PSJ|.

[0120] During the iterative calculation process, if the accuracy meets the requirements, the nonlinear relationship fitting between the burner air volume and the viscous drag coefficient and the inertial drag coefficient can be performed (ie, step S4).

[0121] During the iterative calculation process, if the accuracy does not meet the requirements, the viscous drag coefficient and inertial drag coefficient need to be modified again until the error between PDK and PSJ meets the accuracy requirement of more than 99.9%.

[0122] Implementation 5: Step S4 includes:

[0123] Considering the nonlinear relationship between the burner air volume and the viscous drag coefficient and the inertial drag coefficient, the least squares method is used to fit a quadratic polynomial; the form of the quadratic polynomial is:

[0124] y=a+bx+cx 2

[0125] Where a, b, and c are the coefficients of the constant term, the linear term, and the quadratic term, respectively; x is the mass flow rate, and y is the pressure drop of the burner; y comes from the resistance coefficient dataset PR, and x comes from the operating dataset DO;

[0126] The sum of squared errors is defined as:

[0127]

[0128] By taking the derivative and setting it to 0, we can find the values ​​of a, b, and c:

[0129]

[0130] Finally, the fitted drag coefficient data set FC is formed.

[0131] In this embodiment, the data dimension of the fitted resistance coefficient dataset FC is 3×b, where 3 is the number of multi-item coefficients and b is the number of burner types.

[0132] Implementation method 6: The secondary air volume distribution calculation includes:

[0133] Statistical inlet mass flow rate before the resistance area (i.e. the area where the simplified burner structure is located);

[0134] The resistance coefficient is set by fitting the resistance coefficient data set FC, and the air volume distribution of the secondary air box is numerically simulated.

[0135] The mass flow rates at all burner nozzle outlets are counted to ultimately form a secondary wind box air volume distribution calculation result data set DS.

[0136] In this embodiment, the data dimension of the secondary wind box air volume distribution calculation result data set DS is m×n, where m is the number of all burner outlets and n is the number of typical operating conditions.

[0137] In this embodiment, the resistance coefficient is set by fitting the resistance coefficient data set FC, that is, the viscous resistance coefficient and inertial resistance coefficient in the porous medium region are set by the functional relationship (fitting result) between the air volume and the viscous resistance coefficient and inertial resistance coefficient.

[0138] In this embodiment, a user-defined function (or second user-defined function) in the numerical simulation software is used to set the opening state of the baffle in the wind box.

[0139] Through the user-defined function (UDF) interface provided by the commercial software ANSYS, a DEFINE_ZONE_MOTION type function was customized. The parameters involved are: obj, omega, axis, origin, velocity, time, and dtime. The meanings of the parameters are: function name, baffle opening angle, rotation axis, starting point, rotation speed, calculation time, and iteration interval; the calculation function (the second user-defined function) is expressed as: omega = DEFINE_ZONE_MOTION (axis, origin, velocity, time, dtime). Through the DEFINE_ZONE_MOTION type function named obj, the opening angle of the baffle at different calculation time points is defined, and the function is submitted to the commercial finite element analysis software ANSYS for simulation solution.

[0140] Implementation method 7: Step S6 includes:

[0141] Step 6.1: The secondary air box air volume distribution calculation result dataset DS is combined with the operating data in the operating dataset DO (i.e., secondary air volume SA, secondary air temperature ST, and damper opening state SD) to form basic modeling data, where the secondary air box air volume distribution calculation result dataset DS is the target value;

[0142] Step 6.2: For the basic modeling data, considering the dimensional differences of different parameters, perform max-min normalization on the data:

[0143]

[0144] Among them, x i is the sample value of any parameter in the basic modeling data, x * is the normalized value, x min 、x max are the minimum and maximum values ​​of any parameter sample point respectively;

[0145] Step 6.3: Input the normalized basic modeling data into the ELM neural network calculation model, improve the accuracy of the model calculation by adjusting the parameters, and evaluate the model after denormalizing the calculation results. Based on the evaluation results, select the model with the highest accuracy as the trained online computing machine learning model.

[0146] Embodiment 8: An online calculation device for boiler secondary air box air volume distribution is provided, the device comprising the following modules:

[0147] Module S1: Collect boiler operation data and structural data; generate an operation data set DO based on the operation data; construct a three-dimensional geometric model of the secondary air box and burner based on the structural data;

[0148] Module S2: Based on the operating data set DO and the burner's three-dimensional geometric model, numerical simulation calculations are performed on the pressure drop of a single burner under different air volume conditions to obtain a pressure drop data set PD. The pressure drop statistical boundaries are the burner inlet and outlet. The pressure drop data set PD includes the pressure drops of n typical boiler operating conditions.

[0149] Module S3: Simplify the burner structure in the 3D geometric model of the secondary air box, establish a mapping relationship between the pressure drop in the pressure drop dataset PD and the simplified burner structure, and calculate the numerical correspondence between the pressure drop and the viscous drag coefficient and inertial drag coefficient in the porous media region under the same burner mass flow inlet boundary conditions to form the drag coefficient dataset PR.

[0150] Module S4: Based on the operating data set DO and the drag coefficient data set PR, the least squares method is used to fit the functional relationship between the air volume and the viscous drag coefficient and the inertial drag coefficient by minimizing the sum of squared errors, and the fitted drag coefficient data set FC is obtained;

[0151] Module S5: The simplified burner structure area in the simplified burner structure 3D geometric model of the secondary air box is used as the resistance area. The damper in the air box is set to be open. The data in the running data set DO is used as the boundary condition to perform the secondary air volume distribution calculation and obtain the secondary air box air volume distribution calculation result data set DS.

[0152] Module S6: Using an extreme learning machine to train the numerical mapping relationship between the operating data in the operating data set DO and the secondary air box air volume distribution calculation result data set DS, to obtain a trained online computing machine learning model;

[0153] Module S7: Place the trained online machine learning model in the boiler DCS control system. By inputting real-time operating data, calculate the burner air volume online and output it to the operation screen of the boiler DCS control system for use by operators.

[0154] Embodiment 9: A computer device is provided, comprising: a processor and a memory, wherein the memory is used to store executable instructions of the processor, and the processor is configured to execute any one of the above-mentioned online calculation methods for boiler secondary wind box air volume distribution by executing the executable instructions.

[0155] Implementation 10: A computer storage medium is provided, wherein a computer program is stored in the storage medium. When the computer program is run, any one of the above-mentioned online calculation methods for boiler secondary wind box air volume distribution is executed.

[0156] Embodiment 11: A computer program product is provided, comprising a computer program / instruction, which, when executed by a processor, implements the steps of any one of the above-mentioned methods for online calculation of boiler secondary wind box air volume distribution.

[0157] This embodiment provides a computer device or system, the hardware device of this part is a general model and is not shown in the form of a diagram. The system includes a processor and a memory, wherein the processor and the memory can be connected through a bus or other means. The memory is a non-transient computer-readable storage medium that can be used to store non-transient software programs, non-transient computer executable programs and modules, and corresponding program instructions / modules. The processor executes various functional applications and data processing of the processor by running the non-transient software programs, instructions and modules stored in the memory, so as to realize the data space entity resolution data quality enhancement method in the above method embodiment.

[0158] The memory may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created by the processor, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory may optionally include a memory remotely located relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, an intranet, a mobile communication network, and combinations thereof.

[0159] One or more modules are stored in the memory. When the processor executes, the method steps in the embodiment are executed. In this way, the purpose of the invention can be achieved through the method, device and process of the present invention. The specific details of the above-mentioned computer equipment can be understood by referring to the corresponding descriptions and effects in the embodiment, and will not be repeated here.

[0160] Those skilled in the art will appreciate that all or part of the processes in the above-described embodiments can be implemented by instructing related hardware through a computer program. The program can be stored in a computer-readable storage medium, and when executed, the program can include the processes of the above-described embodiments. The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD). The storage medium can also include a combination of the above-mentioned types of memory.

[0161] The above further describes the technical solution provided by the present invention in detail through several specific embodiments in order to highlight the advantages and benefits of the technical solution provided by the present invention. However, the several specific embodiments described above are not intended to limit the present invention. Any reasonable changes and improvements to the present invention, reasonable combinations of implementation methods and equivalent replacements based on the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. The online calculation method for the air volume distribution of the boiler secondary wind box is characterized by: The method comprises: Step S1: Collect boiler operation data and structural data; form an operation data set DO based on the operation data; construct a three-dimensional geometric model of the secondary wind box and burner based on the structural data; Step S2: Based on the operating data set DO and the burner's three-dimensional geometric model, numerical simulation calculations are performed on the pressure drop of a single burner under different air volume conditions to obtain a pressure drop data set PD, where the pressure drop statistical boundaries are the burner inlet and outlet. The pressure drop data set PD includes the pressure drops of n typical boiler operating conditions. Step S3: Simplify the burner structure in the three-dimensional geometric model of the secondary wind box, establish a mapping relationship between the pressure drop in the pressure drop dataset PD and the simplified burner structure, and calculate the numerical correspondence between the pressure drop and the viscous drag coefficient and inertial drag coefficient in the porous medium region under the same burner mass flow inlet boundary conditions to form a drag coefficient dataset PR; Step S4: Based on the operating data set DO and the drag coefficient data set PR, the least squares method is used to fit the functional relationship between the air volume and the viscous drag coefficient and the inertial drag coefficient respectively by minimizing the sum of squares of the errors, thereby obtaining the fitted drag coefficient data set FC; Step S5: The simplified burner structure area in the three-dimensional geometric model of the secondary air box after the burner structure is simplified is used as the resistance area. The damper in the air box is set to be open. The data in the running data set DO is used as the boundary condition to perform the secondary air volume distribution calculation to obtain the secondary air box air volume distribution calculation result data set DS. Step S6: using an extreme learning machine to train the numerical mapping relationship between the operating data in the operating data set DO and the secondary air box air volume distribution calculation result data set DS, to obtain a trained online computing machine learning model; Step S7: The trained online machine learning model is placed in the boiler DCS control system. By inputting real-time operating data, the burner air volume is calculated online and output to the operation screen of the boiler DCS control system for use by the operator.

2. The online calculation method for boiler secondary air box air volume distribution according to claim 1 is characterized in that: The operating data includes secondary air volume SA, secondary air temperature ST, damper opening state SD and air volume measurement data SM; Run the dataset DO: Among them, [SA i ST i SD i SM i ] represents the operating measurement point data of the typical operating condition i of the boiler, and n is the total number of typical operating conditions of the boiler.

3. The online calculation method for boiler secondary air box air volume distribution according to claim 2 is characterized in that: The step S2 comprises: Finite element analysis is used to mesh the three-dimensional geometric model of the burner and verify its mesh independence. In the burner 3D geometric model after meshing, the operating point data of the typical boiler operating condition i is used. [SA i ST i SD i SM i ] as the boundary condition, the numerical simulation calculation of the pressure drop of a single burner under different air volume conditions is carried out based on the mass conservation, momentum conservation and energy conservation equations; The pressure drop statistical boundaries are the burner inlet and outlet, forming the pressure drop data set PD.

4. The online calculation method for boiler secondary air box air volume distribution according to claim 1 is characterized in that: The step S4 comprises: Considering the nonlinear relationship between the burner air volume and the viscous drag coefficient and the inertial drag coefficient, the least squares method is used to fit a quadratic polynomial; the form of the quadratic polynomial is: y=a+bx+cx 2 Where a, b, and c are the coefficients of the constant term, the linear term, and the quadratic term, respectively; x is the mass flow rate, and y is the pressure drop of the burner; y comes from the resistance coefficient dataset PR, and x comes from the operating dataset DO; The sum of squared errors is defined as: By taking the derivative and setting it to 0, we can find the values ​​of a, b, and c: Finally, the fitted drag coefficient data set FC is formed.

5. The online calculation method for boiler secondary air box air volume distribution according to claim 1 is characterized in that: The secondary air volume distribution calculation includes: Statistical inlet mass flow rate before the resistance area (i.e. the area where the simplified burner structure is located); The resistance coefficient is set by fitting the resistance coefficient data set FC, and the air volume distribution of the secondary air box is numerically simulated. The mass flow rates at all burner nozzle outlets are counted to ultimately form a secondary wind box air volume distribution calculation result data set DS.

6. The online calculation method for boiler secondary air box air volume distribution according to claim 1 is characterized in that: The step S6 comprises: Step 6.1: The secondary air box air volume distribution calculation result dataset DS is combined with the operating data in the operating dataset DO (i.e., secondary air volume SA, secondary air temperature ST, and damper opening state SD) to form basic modeling data, where the secondary air box air volume distribution calculation result dataset DS is the target value; Step 6.2: For the basic modeling data, considering the dimensional differences of different parameters, perform max-min normalization on the data: Among them, x i is the sample value of any parameter in the basic modeling data, x * is the normalized value, x min 、x max are the minimum and maximum values ​​of any parameter sample point respectively; Step 6.3: Input the normalized basic modeling data into the ELM neural network calculation model, improve the accuracy of the model calculation by adjusting the parameters, and evaluate the model after denormalizing the calculation results. Based on the evaluation results, select the model with the highest accuracy as the trained online computing machine learning model.

7. Boiler secondary wind box air volume distribution online calculation device, characterized in that: The device comprises: Module S1: Collect boiler operation data and structural data; generate an operation data set DO based on the operation data; construct a three-dimensional geometric model of the secondary air box and burner based on the structural data; Module S2: Based on the operating data set DO and the burner's three-dimensional geometric model, numerical simulation calculations are performed on the pressure drop of a single burner under different air volume conditions to obtain a pressure drop data set PD. The pressure drop statistical boundaries are the burner inlet and outlet. The pressure drop data set PD includes the pressure drops of n typical boiler operating conditions. Module S3: Simplify the burner structure in the 3D geometric model of the secondary air box, establish a mapping relationship between the pressure drop in the pressure drop dataset PD and the simplified burner structure, and calculate the numerical correspondence between the pressure drop and the viscous drag coefficient and inertial drag coefficient in the porous media region under the same burner mass flow inlet boundary conditions to form the drag coefficient dataset PR. Module S4: Based on the operating data set DO and the drag coefficient data set PR, the least squares method is used to fit the functional relationship between the air volume and the viscous drag coefficient and the inertial drag coefficient by minimizing the sum of squared errors, and the fitted drag coefficient data set FC is obtained; Module S5: The simplified burner structure area in the simplified burner structure 3D geometric model of the secondary air box is used as the resistance area. The damper in the air box is set to be open. The data in the running data set DO is used as the boundary condition to perform the secondary air volume distribution calculation and obtain the secondary air box air volume distribution calculation result data set DS. Module S6: Using an extreme learning machine to train the numerical mapping relationship between the operating data in the operating data set DO and the secondary air box air volume distribution calculation result data set DS, to obtain a trained online computing machine learning model; Module S7: Place the trained online machine learning model in the boiler DCS control system. By inputting real-time operating data, calculate the burner air volume online and output it to the operation screen of the boiler DCS control system for use by operators.

8. A computer device comprising: A processor and a memory, characterized in that the memory is used to store executable instructions of the processor, and the processor is configured to execute the online calculation method for boiler secondary wind box air volume distribution according to any one of claims 1 to 6 by executing the executable instructions.

9. A computer storage medium, characterized in that The storage medium stores a computer program, and when the computer program is run, the online calculation method for boiler secondary wind box air volume distribution according to any one of claims 1 to 6 is executed.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, the steps of the on-line calculation method for boiler secondary wind box air volume distribution according to any one of claims 1 to 6 are implemented.

Images