Boiler combustion instability and heating surface overtemperature state simulation method
By establishing and optimizing the state simulation function, screening and processing boiler data, predicting and controlling the ultra-temperature combustion instability of the boiler heating surface, the common problems of boilers in deep-tuning operation mode are solved, and the operation safety and economicality of the boiler are improved.
Patent Information
- Application Number
- CN202510204325.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-24
- Publication Date
- 2025-06-10
AI Technical Summary
Boilers often face the problem of ultra-temperature combustion instability of the heated surface in deep-tuning operation mode. The prior art is difficult to effectively solve these problems, and there is a lack of relevant state simulation methods.
By screening the sensitive parameters of the boiler, collecting and processing relevant data, and establishing a state simulation function that integrates stateful functions, probability functions and prediction functions to evaluate and optimize the operating status of the boiler, and predict and control the ultra-temperature combustion instability of the heated surface.
It realizes effective prediction and control of ultra-temperature combustion instability of the boiler heating surface, improves the operating safety and economy of the boiler, and reduces equipment failure and maintenance costs.
Smart Images

Figure CN120124466A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of new generation information technology, and particularly relates to a method for simulating the states of unstable boiler combustion and over-temperature of heating surfaces. Background Art
[0002] Under the "dual carbon" strategy, China's energy structure is steadily transforming. Thermal power generation has become the "ballast stone" for large-scale consumption of renewable energy while ensuring energy security, gradually shifting from ensuring basic load to deep flexible peak shaving operation. In the deep regulation (deep adjustment) operation mode of coal-fired power plants, boilers face many challenges. Among them, over-temperature of heating surfaces and unstable combustion are one of the most common problems. These problems not only affect the operation efficiency of the boiler but may also cause equipment damage and safety accidents. In the deep regulation operation mode, rapid load changes can lead to heat accumulation in the combustion chamber or mismatches between fuel supply and air supply, causing a sharp rise in the temperature of the heating surface or incomplete combustion. Improper combustion control, such as an imbalance in the air-coal ratio or unreasonable burner arrangement, can also lead to local high temperatures and unstable combustion. In addition, fouling of the water wall or insufficient water flow will reduce the cooling effect and cause over-temperature of the heating surface; design defects in the combustion chamber or differences in the operating states of burners will result in uneven distribution of heat load, further exacerbating the over-temperature phenomenon. Changes in fuel characteristics, such as different fuel types and particle sizes, and fluctuations in the air intake, such as improper ratios of primary air and secondary air or fluctuations in air pressure, will all affect the stability of combustion.
[0003] To address the problems of unstable boiler combustion and over-temperature of heating surfaces, power plants mostly solve them by means of regular maintenance and repair, equipment transformation, control of coal quality, and optimization of combustion. There are problems such as lag, inflexibility, high costs, and reliance on empirical operation. Currently, there is little research on methods for simulating the states of unstable boiler combustion and over-temperature of heating surfaces, and there are no relevant patents and papers.
[0004] The prior art, such as the Chinese invention patent application with the patent publication number CN117473887A and the name "A Method for Monitoring the Thermal State and Corrosion Prediction of High-Temperature Heating Surfaces of a Waste Incinerator", discloses a method for monitoring the thermal state and corrosion prediction of high-temperature heating surfaces of a waste incinerator, belonging to the technical field of monitoring the thermal state of high-temperature heating surfaces of waste incinerators, including the following processes: obtaining the working medium temperature through thermal calculations based on heat transfer characteristics zoning, obtaining the flue gas temperature and flow rate through three-dimensional numerical simulation, coupling the above two results to obtain the wall temperature of the pipe wall, calculating for typical working conditions, storing the calculation results in a database, and calling the database according to the XGBoost model and calculating to obtain the real-time flue gas temperature, wall temperature, fly ash velocity, and concentration, realizing the monitoring of the thermal condition of high-temperature heating surfaces and implementing over-temperature alarms. The above patent does not have the function of predicting over-temperature of heating surfaces. The monitoring of heating surfaces is only based on traditional thermal calculations and empirical formulas, and the object targeted is only limited to waste incinerators. Summary of the Invention
[0005] In view of the above problems existing in the prior art, a method for simulating the states of unstable boiler combustion and over-temperature of heating surfaces according to the present invention aims to solve the problems of over-temperature of heating surfaces and unstable combustion of boilers in the deep load regulation operation mode by establishing and optimizing a state simulation function and evaluating its implementation effect.
[0006] To achieve the above effects, the technical solution adopted by the present invention is as follows:
[0007] A method for simulating the states of unstable boiler combustion and over-temperature of heating surfaces includes the following steps:
[0008] Step (1): Screening of boiler sensitive parameters: The boiler sensitive parameters include parameters related to over-temperature of boiler heating surfaces and parameters related to unstable boiler combustion. Among them, the parameters related to over-temperature of boiler heating surfaces include discrete operation control parameters and continuous operation control parameters;
[0009] Step (2): Boiler data acquisition: Boiler data acquisition includes acquisition of boiler unstable combustion fault data, acquisition of boiler over-temperature fault data, and acquisition of boiler normal operation data;
[0010] Step (3): Boiler data processing: Based on the historical operation data of the unit, statistical and machine learning methods are used to process the collected operation parameters;
[0011] Step (4): Establishing a state simulation function: The state simulation function incorporates a state function, a probability function, and a prediction function;
[0012] Step (5): Evaluating the implementation effect.
[0013] Further, the discrete operation control parameters in step (1) are as follows:
[0014] ① Flue gas damper regulating valve position: used to control the flue gas flow path;
[0015] ② Secondary air regulating valve position: used to adjust the amount of secondary air entering the combustion chamber;
[0016] ③ Gas air regulating valve position: used to adjust the amount of gas air entering the burner.
[0017] Further, the continuous operation control parameters in step (1) are as follows:
[0018] ① Furnace air volume: affecting the oxygen supply in the combustion chamber;
[0019] ② Economizer inlet flow rate: affecting the water flow rate entering the economizer and the heat exchange efficiency;
[0020] ③ Steam temperatures at the inlets of the first and second stage attemperators: affecting the cooling effect of the attemperators;
[0021] ④ Main steam temperature: directly reflects the temperature of the steam at the boiler outlet.
[0022] Further, the parameters related to unstable boiler combustion specifically include:
[0023] ① Furnace wind pressure: affects the pressure distribution in the combustion chamber;
[0024] ② Coal flame detection intensity: reflects the combustion intensity of pulverized coal in the combustion chamber;
[0025] ③ Pulverized coal pipeline air velocity: affects the distribution and combustion rate of pulverized coal in the combustion chamber;
[0026] ④ Combustion system operation control parameters: include the working state of the burners, air-coal ratio, fuel input, fuel / air ratio, primary air damper ratio, secondary air damper opening.
[0027] Further, in the step (2),
[0028] The boiler combustion instability fault data includes the flame detection intensity signals of 8 burners on each layer of the boiler A - F layer burners, the coal feeding amounts of 6 coal mills, the coal mill current signals, the primary air flow rates at the inlets of each coal mill, the flow velocities of the pulverized coal pipelines in front of each burner and the pulverized coal concentrations in the pipelines, the air flow rates of the secondary air on each layer of the boiler, the air flow rates of the overfire air, the opening degrees of the secondary air dampers on each layer, the opening degrees of the tertiary air dampers of each burner and the opening degrees of the center air damper, the opening degree of the overfire air damper, the oxygen content at the tail of the furnace, the total air volume, the total coal amount, the coal quality information, and the temperature of the secondary air at the outlet of the air preheater;
[0029] The boiler over-temperature fault data mainly includes the wall temperatures of each tube of the spiral water wall, the wall temperatures of each tube of the vertical water wall, the wall temperatures of each tube of the platen superheater, the wall temperatures of each tube of the high-temperature superheater, the wall temperatures of each tube of the low-temperature superheater, the wall temperatures of each tube of the high-temperature reheater, the wall temperatures of each tube of the low-temperature reheater, the outlet pressure of the high-temperature superheater, the boiler load, the boiler feed water flow rate, the boiler feed water temperature, the total air volume, the total fuel amount, the main steam parameters, the desuperheating water flow rate of the superheater, the outlet temperature of the high-temperature superheater, the inlet / outlet temperatures of the platen superheater, the temperature of the pulverized coal-air mixture at the outlet of the coal mill;
[0030] The boiler normal operation data includes the boiler load, load change rate, steam temperature, pressure, air supply volume, coal feeding amount, coal quality, flue gas temperature, oxygen content, carbon content in fly ash / slag, and historical data of the burner operation mode, primary air and secondary air parameters.
[0031] Further, the processing in the step (3) specifically includes the following contents:
[0032] ① Operating condition identification: By analyzing historical data, the operating parameter characteristics under different operating conditions are identified. The operating data of the boiler under different loads are different. As the load increases, the operating parameters obviously change linearly.
[0033] ② Abnormal data diagnosis: detect and eliminate abnormal values in the data. Using the standard deviation method, data exceeding ±3 times the standard deviation of the data mean is considered abnormal data;
[0034] ③ Missing data filling: For missing data, interpolation or prediction methods are used to fill it;
[0035] ④Data cleaning: remove irrelevant or redundant data.
[0036] Furthermore, there are errors in the historical data of the power plant that need to be processed, and the errors that need to be processed include systematic errors and random errors. The systematic errors are caused by reading errors, measurement method errors or measurement instrument errors; the random errors are fluctuations in measurement results caused by external interference;
[0037] Data preprocessing is performed for system errors. The data preprocessing method is as follows:
[0038] Let the sample data be y 1 ,y 2 , ..., y n , the average value is The deviation is The standard deviation σ is calculated according to the Bessel formula:
[0039]
[0040] When a sample data deviation v k (1≤k≤n) satisfies |v k |>3σ, then it is considered that the data y k It is unreasonable and should be eliminated, where n is the number of data, i represents a certain data, and i is variable.
[0041] For boiler historical data, the preprocessed data input matrix uses row vector samples. The number of columns in the input matrix represents the total number of model inputs, while the row vector represents each collected data sample;
[0042] In order to eliminate the random errors in the measurement data, a smoothing algorithm is used as follows:
[0043] Assume that some sample data is:
[0044] {x 1 , x 2 , …, x n}, x i =s i +ni ;
[0045] where \(x\) represents sample data, \(s\) i is the true value, \(n\) i is the noise. In the above formula, \(n\) is the number of data, \(i\) represents a certain data, and \(i\) is variable. Then
[0046]
[0047] In the formula, \(\{a\) r \} is the weight, satisfying \(y\) i is the processed data, \(q\) is the average value of a section of distance, \(\Delta\) is the step size of the window, \(r\) represents the value within the distance from \(-q\) to \(q\), and \(a\) r represents the corresponding weight within the distance \(r\), \(x\) i+r is a certain data considering the change of the distance \(r\), \(s\) i+r is a certain true value considering the distance \(r\), \(n\) i+r is a certain error value considering the distance \(r\).
[0048] Furthermore, in the step (4):
[0049] ① State function: A state function that includes all relevant measured point data, used to describe the current state of the system, including the selected boiler sensitive parameters;
[0050] ② Probability function: A probability function of the influence degree of measured point data on faults, used to evaluate the influence degree of each parameter on overheating of the heating surface and unstable combustion;
[0051] ③ Prediction function: A prediction function of the change trend of measured point data after performing a certain operation, used to predict the influence of different operations on the system state;
[0052] The specific form of the state simulation function is as follows
[0053]
[0054] In the formula, \(r\) t is the target variable at time \(t\); \(s\) t is the current system state; \(a\) t is a discrete decision variable, selected from the finite decision set \(A\) through the policy function \(\pi(\cdot)\), and \(t_0\) represents the initial time;
[0055] \(R(\cdot)\) is the objective function at time \(t\), reflecting the combustion safety;
[0056] \(T(\cdot)\) is the state transition function, reflecting the relationship of system state changes;
[0057] \(C(\cdot)\) is the state trajectory inequality constraint, where \(k\) is an indefinite value, and the range is from 1 to \(m\); \(m\) represents the number of digits;
[0058] φ(·) is the initial value equality constraint, where l is an indefinite value ranging from 1 to m; m represents the number of digits;
[0059] h(·) is the action equation, where i is an indefinite value ranging from 1 to p; p represents the number of digits; i represents any data in a set of data;
[0060] g(·) is the inequality constraint function, where j is an indefinite value ranging from 1 to q; q represents the number of digits.
[0061] Furthermore, in the step (5), the specific content of evaluating the implementation effect is:
[0062] For the input parameters: the wall temperatures of each tube of the spiral water wall, the wall temperatures of each tube of the vertical water wall, the wall temperatures of each tube of the platen superheater, the wall temperatures of each tube of the high-temperature superheater, the wall temperatures of each tube of the low-temperature superheater, the wall temperatures of each tube of the high-temperature reheater, the wall temperatures of each tube of the low-temperature reheater, the outlet temperature of the high-temperature superheater, the boiler load, and the boiler feed water flow rate, the output parameter is the predicted temperature value, and the temperature control accuracy is evaluated by comparing the difference between the predicted value and the actual value;
[0063] For the input parameters: the flame detector intensity signals of 8 burners in each layer of the A - F layer burners of the boiler, a statistical model based on the flame detector intensity signals is established, and the mean and variance of the flame detector intensity signals are calculated. If the mean of the flame detector intensity signals decreases or the variance increases, it indicates that the flame detector intensity is unstable or decreasing;
[0064] A logistic regression model or a support vector machine model based on the wall temperature is established to judge whether the heating surface is overheated. The input parameters are the wall temperatures of each tube of the spiral water wall, the wall temperatures of each tube of the vertical water wall, the wall temperatures of each tube of the platen superheater, the wall temperatures of each tube of the high-temperature superheater, the wall temperatures of each tube of the low-temperature superheater, the wall temperatures of each tube of the high-temperature reheater, and the wall temperatures of each tube of the low-temperature reheater, and the output is the overheating judgment result;
[0065] A multiple linear regression model or a neural network model is established. The input parameters are the boiler load, the load change rate, the steam temperature, the pressure, the air supply volume, the coal feeding amount, the coal quality, the flue gas temperature, the oxygen content, the carbon content in fly ash / slag, and the operation mode of the burners, and the output parameters are the evaluation indexes of the thermal efficiency or the combustion stability;
[0066] A multiple linear regression model or a neural network model based on the wind parameters is established, and the output is the evaluation index of the uniformity.
[0067] The advantages of this application are:
[0068] 1. This application first screens and collects boiler operation data, and then processes the data using statistical and machine learning methods. It has strong capabilities in processing large-scale data, improves the quality of the data, and further enhances the accuracy of the model.
[0069] 2. This application provides a method for simulating the states of unstable boiler combustion and over-temperature of heating surfaces, and proposes a process for establishing and optimizing the state simulation function. It can effectively predict and control the problems of over-temperature of heating surfaces and unstable combustion during deep regulation of the boiler, thereby improving the operating safety and economy of the boiler.
[0070] 3. The solution of this application uses big data and neural network methods to make the results more credible and persuasive, and has the ability of monitoring and early warning compared with the prior art. BRIEF DESCRIPTION OF THE DRAWINGS
[0071] Figure 1 is a flow chart of the present invention.
[0072] Figure 2 is a state transition function.
[0073] Figure 3 is the boiler structure and burner layout. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0074] To make the objectives, technical solutions and advantages of the embodiments of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are some but not all of the embodiments of this application. Usually, the components of the embodiments of this application described and shown in the drawings here can be arranged and designed in various different configurations.
[0075] Therefore, the following detailed description of the embodiments of this application provided in the drawings is not intended to limit the scope of this application which is claimed, but merely represents selected embodiments of this application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in this application without creative efforts fall within the scope of protection of this application.
[0076] It should be noted that: similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.
[0077] In the description of the present application, it should be noted that the orientation or positional relationship indicated by terms such as "upper", "vertical", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship in which the product of this application is usually placed during use, or the orientation or positional relationship commonly understood by those skilled in the art. It is only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present application. In addition, terms such as "first", "second", etc. are only used for differential description and should not be construed as indicating or implying relative importance.
[0078] In the description of the present application, it should also be noted that unless otherwise clearly specified and limited, the terms "set", "installed", "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to specific circumstances.
[0079] Embodiment 1
[0080] A method for simulating the state of unstable boiler combustion and overheating of heating surfaces includes the following steps:
[0081] Step (1): Screening of boiler sensitive parameters: According to the unit system structure, the boiler sensitive parameters include parameters related to overheating of boiler heating surfaces and parameters related to unstable boiler combustion. Among them, the parameters related to overheating of boiler heating surfaces include discrete operation control parameters and continuous operation control parameters;
[0082] Step (2): Boiler data acquisition: Boiler data acquisition includes acquisition of boiler unstable combustion fault data, acquisition of boiler overheating fault data, and acquisition of boiler normal operation data;
[0083] Step (3): Boiler data processing: Based on the historical operation data of the unit, statistical and machine learning methods are used to process the collected operation parameters; knowledge of statistics is involved in the processes of data screening, processing, and analysis, and machine learning methods are involved in establishing the state simulation function and evaluating the implementation effect.
[0084] Step (4): Establishing a state simulation function: The state simulation function incorporates a state function, a probability function, and a prediction function;
[0085] Step (5): Evaluating the implementation effect.
[0086] Embodiment 2
[0087] A method for simulating the state of unstable boiler combustion and overheating of heating surfaces includes the following steps:
[0088] Step (1): Screening of boiler sensitive parameters: According to the unit system structure, the boiler sensitive parameters include parameters related to overheating of boiler heating surfaces and parameters related to unstable boiler combustion. Among them, the parameters related to overheating of boiler heating surfaces include discrete operation control parameters and continuous operation control parameters;
[0089] Step (2): Boiler data acquisition: Boiler data acquisition includes acquisition of boiler unstable combustion fault data, acquisition of boiler overheating fault data, and acquisition of boiler normal operation data;
[0090] Step (3): Boiler data processing: Based on the historical operation data of the unit, statistical and machine learning methods are used to process the collected operation parameters; knowledge of statistics is involved in the processes of data screening, processing, and analysis, and machine learning methods are involved in establishing the state simulation function and evaluating the implementation effect.
[0091] Step (4): Establishing the state simulation function: The state simulation function incorporates a state function, a probability function, and a prediction function;
[0092] Step (5): Evaluating the implementation effect.
[0093] The discrete operation control parameters in step (1) are:
[0094] ① Flue gas damper regulating valve position: Used to control the flue gas flow path, affecting the flue gas temperature and flow rate, and thus affecting the temperature distribution of the heating surface.
[0095] ② Secondary air regulating valve position: Used to adjust the amount of secondary air entering the combustion chamber, affecting the combustion efficiency and temperature distribution, and further affecting the temperature of the heating surface.
[0096] ③ Gas-air regulating valve position: Used to adjust the gas-air volume entering the burner, affecting the combustion stability and temperature, and thus affecting the temperature of the heating surface.
[0097] Furthermore, the continuous operation control parameters in step (1) are:
[0098] ① Furnace air volume: Affects the oxygen supply in the combustion chamber, has an important impact on the combustion process and temperature distribution, and further affects the temperature of the heating surface.
[0099] ② Economizer inlet flow rate: Affects the water flow rate entering the economizer, affects the heat exchange efficiency, and thus affects the temperature of the heating surface.
[0100] ③ Inlet steam temperatures of the first and second stage desuperheaters: Affect the cooling effect of the desuperheater, and thus affect the main steam temperature and the temperature of the heating surface.
[0101] ④ Main steam temperature: It directly reflects the temperature of the steam at the boiler outlet and is an important indicator for judging whether the heating surface is overheated.
[0102] Further, the parameters related to unstable boiler combustion specifically include:
[0103] ① Furnace wind pressure: It affects the pressure distribution in the combustion chamber and has an important impact on the stability of the combustion process.
[0104] ② Coal flame detection intensity: It reflects the combustion intensity of pulverized coal in the combustion chamber and has a direct impact on combustion stability.
[0105] ③ Pulverized coal pipeline wind speed: It affects the distribution and combustion rate of pulverized coal in the combustion chamber and has an important impact on combustion stability.
[0106] ④ Combustion system operation control parameters: including the working state of burners, air-coal ratio, fuel input, fuel / air ratio, primary air damper ratio, secondary air damper opening. These parameters directly affect the stability of the combustion process.
[0107] In step (2),
[0108] The boiler combustion instability fault data includes the flame detection intensity signals of 8 burners on each layer of the A - F layer burners of the boiler, the coal feeding amounts of 6 coal mills, the coal mill current signals, the primary air flow rates at the inlets of each coal mill, the flow velocities of pulverized coal pipelines in front of each burner and the pulverized coal concentrations in the pipelines, the air flow rates of secondary air on each layer of the boiler, the air flow rates of overfire air, the opening degrees of secondary air dampers on each layer, the opening degrees of tertiary air dampers of each burner and the opening degrees of center air dampers, overfire air damper opening degrees, oxygen content at the tail of the furnace, total air volume, total coal amount, coal quality information, and the temperature of secondary air at the outlet of the air preheater, etc.
[0109] The boiler overheating fault data mainly includes the wall temperatures of each tube of the spiral water - cooled wall, the wall temperatures of each tube of the vertical water - cooled wall, the wall temperatures of each tube of the platen superheater, the wall temperatures of each tube of the high - temperature superheater, the wall temperatures of each tube of the low - temperature superheater, the wall temperatures of each tube of the high - temperature reheater, the wall temperatures of each tube of the low - temperature reheater, the outlet pressure of the high - temperature superheater, boiler load, boiler feed water flow rate, boiler feed water temperature, total air volume, total fuel amount, main steam parameters (pressure, temperature, flow rate), desuperheating water flow rate of the superheater, outlet temperature of the high - temperature superheater, inlet / outlet temperatures of the platen superheater, temperature of the pulverized coal - air mixture at the outlet of the coal mill, etc.
[0110] The normal operation data of the boiler comprehensively considers the historical operation data of the boiler under different loads, different variable load rates, different grinding group commissioning, different combustion organization modes and burner operation modes. The collected data list also covers the key operating parameters in the boiler heating surface overtemperature and combustion instability fault data list, mainly including boiler load, variable load rate (calculated value), steam temperature, pressure, air supply, coal supply, coal quality, exhaust gas temperature, oxygen content, fly ash / slag carbon content, burner commissioning mode, primary air and secondary air parameters (flow, temperature, pressure, damper opening, etc.) historical data.
[0111] The processing in step (3) specifically includes the following contents:
[0112] ① Operating condition identification: By analyzing historical data, the operating parameter characteristics under different operating conditions are identified to provide a basis for subsequent fault diagnosis and prediction. The operating data of the boiler under different loads are different. As the load increases, the operating parameters obviously change (increase) in an approximately linear manner.
[0113] ② Abnormal data diagnosis: Detect and remove abnormal values in the data to ensure the accuracy and reliability of the data. Using the standard deviation method, data exceeding ±3 times the standard deviation of the data mean is considered abnormal data.
[0114] ③ Missing data filling: For missing data, interpolation or prediction methods are used to fill it in to ensure data integrity.
[0115] ④Data cleaning: remove irrelevant or redundant data, retain data that is important for research, and improve data quality. Experience can be used to determine and obtain the power plant operation platform without warning data set.
[0116] For a specific power plant, there are errors in the historical data of the power plant that need to be processed. The errors here are objective, and there are errors in the processed data. The errors that need to be processed include systematic errors and random errors. The systematic errors are caused by reading errors, measurement method errors, or measurement instrument errors; the random errors are mainly caused by fluctuations in measurement results due to external interference and other reasons;
[0117] In order to deal with system errors, data preprocessing is required at the beginning, and this step is usually carried out simultaneously with data cleaning. System errors are less likely to occur, but their presence will deteriorate the data quality and thus affect the final results. When preprocessing data, it is necessary to first remove the bad data. The data preprocessing method is as follows:
[0118] Let the sample data be y 1 ,y 2 , ..., y n , the average value is The deviation is The standard deviation σ is calculated according to Bessel's formula:
[0119]
[0120] When the deviation v of a certain sample data k (1 ≤ k ≤ n) satisfies |v k | > 3σ, then the data y is considered k unreasonable and should be excluded. Here, n is the number of data, i represents a certain data, and i is variable.
[0121] In actual data processing, after obtaining the historical data of the boiler, the pre - processed data input matrix uses row - vector samples. That is, the number of columns of the input matrix represents the total number of model inputs, and the row vector represents each collected data sample. The error - processing process is to identify and exclude the problematic row vectors.
[0122] For the random errors in the measurement data, a smoothing algorithm is used to eliminate them. Among them, the linear smoothing algorithm is more commonly used in smoothing data, and the method is as follows:
[0123] Suppose a certain sample data is:
[0124] {x 1 , x 2 , …, x n}, x i = s i + n i ;
[0125] where x represents the sample data, s i is the true value, n i is the noise. In the above formula, n is the number of data, i represents a certain data, and i is variable. The true value is the absolutely correct value, and the noise is the magnitude of the error value. Then
[0126]
[0127] In the formula, {a r} is the weight, satisfying y i is the processed data, q is the average value of a section of distance, is the step size of the window, r represents the value within the distance from - q to q, a r represents the weight corresponding to the r - distance, x i+r is a certain data after considering the change of the r - distance, s i+r is a certain true value after considering the r - distance, n i+r is a certain error value after considering the r - distance.
[0128] For the data that cannot be processed by the above method, it is necessary to make judgments and selections in combination with the operating characteristics and experience of the boiler.
[0129] In step (4), the establishment of the state simulation function is the core part of the implementation steps, and it includes three main parts:
[0130] ① State function: The state function that contains all relevant measured point data is used to describe the current state of the system. These data include the boiler sensitive parameters selected above, such as the flue gas damper regulating valve position, furnace air volume, etc.
[0131] ② Probability function: The probability function of the influence degree of the measured point data on the fault is used to evaluate the influence degree of each parameter on the overheating of the heating surface and the unstable combustion. Through the probability function, it is possible to identify which parameters have the greatest influence on the fault, so as to take targeted control measures.
[0132] ③ Prediction function: The prediction function of the change trend of the measured point data after performing a certain operation is used to predict the influence of different operations on the system state. Through the prediction function, we can anticipate the system changes after the operation in advance, so as to optimize the operation strategy and avoid the occurrence of faults.
[0133] The state simulation function is an integral whole. The state simulation function contains the state function, the probability function, and the prediction function. The three are not independent and need to be combined into an integral whole. The specific form is as follows
[0134]
[0135] In the formula, r t is the target variable at time t; s t is the current system state; a t is the discrete decision variable, which is selected from the finite decision set A through the policy function π(·), and t0 represents the initial time;
[0136] R(·) is the objective function at time t, which reflects the combustion safety (no overheating and stable combustion);
[0137] T(·) is the state transition function, which reflects the system state change relationship;
[0138] C(·) is the state trajectory inequality constraint, where k is an indefinite value, and the range is from 1 to m; m represents the number of digits;
[0139] φ(·) is the initial value equality constraint, where l is an indefinite value, and the range is from 1 to m; m represents the number of digits;
[0140] h(·) is the action equality, where i is an indefinite value, and the range is from 1 to p; p represents the number of digits; i represents any data in a set of data;
[0141] g(·) is the inequality constraint function, where j is an indefinite value, and the range is from 1 to q; q represents the number of digits.
[0142] By establishing and optimizing the state simulation function, we can more accurately predict and control the problems of overheating of the heating surface and combustion instability during the deep regulation of the boiler, thereby improving the operating safety and economy of the boiler. Figure 2 Shows the specific form of the state transition function T(·), which helps us better understand the law of change of the system state. Similar to the Monte Carlo method, it shows the state transition process of a time series data, which is particularly suitable for the scenario of long short-term memory network (LSTM). The figure shows the prediction of the next state based on the current state for two states (state k and state k+1). Each state contains four features (Lk, Sk, Mk, Tk). The time axis at the top of the figure represents different time points of the time series, from tk to tk+Np+1. State k is the state at the current time step and contains four features; state k+1 is the state at the next time step and also contains four features. In this way, LSTM can effectively utilize historical information to predict future states or outputs when processing time series data. At each time step, the model updates the state based on the current state and input data and predicts the state at the next time step.
[0143] In step (5), the specific content of evaluating the implementation effect is:
[0144] A multiple linear regression model can be used. For the input parameters: the wall temperatures of each tube of the spiral water wall, the wall temperatures of each tube of the vertical water wall, the wall temperatures of each tube of the platen superheater, the wall temperatures of each tube of the high-temperature superheater, the wall temperatures of each tube of the low-temperature superheater, the wall temperatures of each tube of the high-temperature reheater, the wall temperatures of each tube of the low-temperature reheater, the outlet temperature of the high-temperature superheater, the boiler load, and the boiler feed water flow, and the output parameter is the predicted temperature value. The temperature control accuracy can be evaluated by comparing the difference between the predicted value and the actual value; a neural network model can also be adopted to utilize its powerful nonlinear fitting ability to more accurately capture the relationship between temperature and various influencing factors.
[0145] For the input parameters: the flame detector intensity signals of 8 burners in each layer of the A-F layer burners of the boiler, a statistical model based on the flame detector intensity signals can be established. For example, calculate the mean and variance of the flame detector intensity signals. If the mean of the flame detector intensity signals decreases or the variance increases, it may indicate that the flame detector intensity is unstable or decreasing; a time series analysis model can also be adopted to perform trend analysis and anomaly detection on the flame detector intensity signals to timely detect changes in the flame detector intensity.
[0146] A logical regression model or a support vector machine model based on the tube wall temperature can be established to judge whether the heating surface is overheated. The input parameters are the tube wall temperatures of each spiral water wall, each vertical water wall, each platen superheater, each high-temperature superheater, each low-temperature superheater, each high-temperature reheater, and each low-temperature reheater, and the output is the overheating judgment result (yes / no). A decision tree model can also be used to classify whether it is overheated by setting a temperature threshold.
[0147] A multiple linear regression model or a neural network model can be established. The input parameters are the boiler load, variable load rate (calculated value), steam temperature, pressure, air supply volume, coal feeding volume, coal quality, flue gas temperature, oxygen content, carbon content in fly ash / slag, and the operation mode of burners, and the output parameters are the evaluation indexes of thermal efficiency or combustion stability; A fuzzy comprehensive evaluation model can also be used to convert multiple influencing factors into fuzzy sets and obtain the comprehensive evaluation results of thermal efficiency or combustion stability through fuzzy operations.
[0148] A multiple linear regression model or a neural network model based on wind parameters can be established to evaluate the uniformity of the wind. The inputs of the model are the primary air and secondary air parameters (flow rate, temperature, pressure, damper opening, etc.), and the output is the evaluation index of uniformity. A clustering analysis model can also be used to classify samples with similar wind parameters into one category and evaluate the uniformity of the wind by comparing the differences between different categories.
[0149] Generally speaking, the evaluation of the implementation effect includes whether the temperature control accuracy is improved, whether the flame detection intensity is decreased, whether the heating surface is overheated, the thermal efficiency, combustion stability, and whether the uniformity is improved, etc. The implementation effect is evaluated from the aspects of solving the overheating problem of the heating surface, solving the combustion instability problem, economic benefits, environmental benefits, and user satisfaction.
[0150] Embodiment 3
[0151] As Figure 1 shown, a method for simulating the state of boiler combustion instability and heating surface overheating includes the following steps:
[0152] Step (1): Screening of boiler sensitive parameters;
[0153] Step (2): Boiler data collection;
[0154] Step (3): Boiler data processing;
[0155] Step (4): Establishing a state simulation function;
[0156] Step (5): Evaluating the implementation effect.
[0157] In step (1), as Figure 3As shown in the figure, a 1000MW ultra-supercritical parameter, variable pressure once-through boiler, front and rear wall opposed firing, solid slagging, single furnace, once reheat, balanced draft, all-steel structure (with tight enclosure), and fully suspended π-shaped structure boiler in a certain power plant is selected, adopting the front and rear wall opposed firing method. According to the unit system structure, the parameters related to the over-temperature of the boiler heating surface are found. These parameters can be divided into two categories: discrete operation control parameters and continuous operation control parameters.
[0158] The discrete operation control parameters are:
[0159] ① The position of the flue gas damper regulating valve: Controls the flue gas flow path, affects the flue gas temperature and flow rate, and thus affects the temperature distribution of the heating surface.
[0160] ② The position of the secondary air regulating valve: Regulates the amount of secondary air entering the combustion chamber, affects the combustion efficiency and temperature distribution, and further affects the temperature of the heating surface.
[0161] ③ The position of the gas-air regulating valve: Regulates the gas-air volume entering the burner, affects the combustion stability and temperature, and thus affects the temperature of the heating surface.
[0162] The continuous operation control parameters are:
[0163] ① The furnace air volume: Affects the oxygen supply in the combustion chamber, has an important impact on the combustion process and temperature distribution, and further affects the temperature of the heating surface.
[0164] ② The flow rate at the inlet of the economizer: Affects the water flow rate entering the economizer, affects the heat exchange efficiency, and thus affects the temperature of the heating surface.
[0165] ③ The steam temperature at the inlets of the first and second stage desuperheaters: Affects the cooling effect of the desuperheater, and thus affects the main steam temperature and the temperature of the heating surface.
[0166] ④ The main steam temperature: Directly reflects the temperature of the steam at the boiler outlet and is an important indicator for judging whether the heating surface is over-temperature.
[0167] Secondly, according to the unit system structure, the main parameters related to the unstable combustion of the boiler also need to be found:
[0168] ① The furnace wind pressure: Affects the pressure distribution in the combustion chamber and has an important impact on the stability of the combustion process.
[0169] ② The intensity of the coal flame detector: Reflects the combustion intensity of pulverized coal in the combustion chamber and has a direct impact on the combustion stability.
[0170] ③ The wind speed in the pulverized coal pipeline: Affects the distribution and combustion rate of pulverized coal in the combustion chamber and has an important impact on the combustion stability.
[0171] ④Main operating control parameters of the combustion system: including the working state of the burner, the air-coal ratio, etc. These parameters directly affect the stability of the combustion process.
[0172] In step (2), when collecting boiler data, select data for a continuous time period, such as from June to July, which should cover the typical load boiler operation data between 30% and 100% of the boiler load section. The collected typical historical data of the boiler provides a basis for the establishment of the state function. However, the establishment of the model is restricted by reasons such as boiler shutdown, maintenance, and transformation. Further collection and analysis of the relevant operation data of the boiler can continuously optimize the relevant model and improve the model accuracy.
[0173] In step (3), data processing is the basis for the analysis of boiler combustion faults and the optimization of modeling work, and has a crucial impact on the accuracy and effectiveness of the model. Affected by factors such as instrument detection and harsh on-site working conditions, the boiler operation data fluctuates greatly, and due to the real-time changes of various boiler parameters, it is necessary to screen out the boiler operating conditions and their operation data with stable load sections and no abnormalities in various boiler parameters, so as to ensure that the modeling data is representative and reliable.
[0174] In step (4), the establishment of the state simulation function is the core part of the implementation step, which includes three main parts:
[0175] ①State function: A state function that contains all relevant measured point data, used to describe the current state of the system. These data include the sensitive parameters selected above, such as the position of the flue gas damper regulating valve, the furnace air volume, etc.
[0176] ②Probability function: A probability function of the influence degree of the measured point data on the fault, used to evaluate the influence degree of each parameter on the overheating of the heating surface and the instability of combustion. Through the probability function, we can identify which parameters have the greatest impact on the fault, so as to take targeted control measures.
[0177] ③Prediction function: A prediction function of the change trend of the measured point data after performing a certain operation, used to predict the influence of different operations on the system state. Through the prediction function, we can anticipate the system changes after the operation in advance, so as to optimize the operation strategy and avoid faults.
[0178] The specific form of the state simulation function is as follows:
[0179]
[0180] In the formula, r t is the target variable at time t; r t is the current system state; a tis a discrete decision variable, which is selected from a finite decision set A through the policy function π(·); R(·) is the objective function at time t, reflecting combustion safety (no overheating and stable combustion); T(·) is the state transition function, reflecting the relationship of system state changes; C(·), φ(·), h(·) and g(·) are the state trajectory inequality constraint, initial value equality constraint, action equality and inequality constraint functions respectively. By establishing and optimizing the state simulation function, we can more accurately predict and control the overheating of the heating surface and unstable combustion problems during the deep adjustment of the boiler, thereby improving the operating safety and economy of the boiler. Figure 2 Shows the specific form of the state transition function T(·), which helps us better understand the change law of the system state.
[0181] In step (5), evaluate the implementation effect, and evaluate from the aspects of solving the overheating problem of the heating surface, solving the unstable combustion problem, and economic benefits, environmental benefits, and user satisfaction.
[0182] i) Solution to the overheating problem of the heating surface
[0183] ①Improved temperature control accuracy: Through the state simulation function, we can more accurately predict and control the temperature of the heating surface. In actual operation, the system can timely adjust parameters such as the position of the flue gas damper regulating valve and the position of the secondary air regulating valve, effectively preventing the occurrence of overheating of the heating surface. The experimental results show that the temperature control accuracy has increased by about 20%.
[0184] ②Extended equipment life: The stable control of the heating surface temperature reduces the thermal stress and creep damage of the material, extending the service life of the equipment. It is estimated that the average life of the equipment has been extended by about 15%.
[0185] ③Improved safety: The reduction of the overheating problem significantly reduces the risk of equipment failure and safety accidents. During the operation cycle after the implementation of the algorithm, no equipment failure or safety accident caused by overheating occurred.
[0186] ④Improved thermal efficiency: The stable control of the heating surface temperature makes the heat transfer more uniform, improving the thermal efficiency. The experimental data show that the thermal efficiency has increased by about 5%.
[0187] ii) Solution to the unstable combustion problem
[0188] ①Improved combustion efficiency: By optimizing parameters such as furnace wind pressure, intensity of coal flame detector, and pulverized coal pipeline air velocity, the combustion process becomes more stable and efficient. The experimental data show that the combustion efficiency has increased by about 10%, and the fuel consumption has decreased significantly.
[0189] ②Emission reduction: The improvement of combustion efficiency and the stability of the combustion process reduce the generation of unburned products, thereby reducing pollutant emissions. The NOx emissions are reduced by about 15%, and the CO emissions are reduced by about 20%.
[0190] ③Enhanced operation stability: After the implementation of the algorithm, the fluctuations in the combustion process are significantly reduced, and the system operates more smoothly. The operating personnel feedback that the operation difficulty is reduced, the system response speed is faster, and the overall operation stability is significantly enhanced.
[0191] ④Improved combustion uniformity: By optimizing the burner layout and the air-coal ratio, the combustion uniformity is significantly improved, reducing the phenomena of local high temperature and incomplete combustion.
[0192] iii)Improved economic benefits
[0193] ①Reduced operation cost: The improvement of combustion efficiency and the reduction of fuel consumption directly reduce the operation cost. It is estimated that after the implementation of the algorithm, each boiler can save about 10% of the fuel cost per year.
[0194] ②Reduced maintenance cost: The extension of equipment life and the reduction of failure rate reduce the frequency of maintenance and repair, further reducing the maintenance cost. According to statistics, the maintenance cost is reduced by about 20%.
[0195] ③Reduced downtime: Due to the reduction of equipment failures and safety accidents, the downtime of the boiler is significantly shortened, improving the overall operation efficiency of the power plant. The downtime is reduced by about 30%, and the power generation increases.
[0196] ④Significant economic benefits: Considering the reduction of fuel cost, maintenance cost and downtime comprehensively, the economic benefits after the implementation of the algorithm are significantly improved. It is estimated that each boiler can increase the economic benefits by about 15% per year.
[0197] (iv)Significant environmental benefits
[0198] ①Reduced carbon emissions: The improvement of combustion efficiency and the reduction of fuel consumption directly reduce the CO 2 emissions, having a positive impact on environmental protection. It is estimated that each boiler can reduce the CO 2 emissions by about 10% per year.
[0199] ②Pollutant reduction: The reduction of NOx and CO emissions not only improves the air quality but also meets the requirements of environmental protection regulations, enhancing the social responsibility image of the enterprise.
[0200] ③Environmentally friendly: By reducing pollutant emissions and improving energy utilization efficiency, the boiler operation is more environmentally friendly, contributing to the achievement of sustainable development goals.
[0201] (v)Improved user satisfaction
[0202] ① Simple operation: After the implementation of the algorithm, the automation level of the system is improved, the workload of the operators is reduced, and the simplicity of operation is significantly enhanced. The operators feedback that the system operation is more intuitive and convenient.
[0203] ② Fault warning: Through the state simulation function, the system can give early warning of potential faults, helping the operators to take timely measures to avoid the occurrence of faults. The accuracy rate of the fault warning reaches more than 85%.
[0204] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, and the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.
[0205] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.
[0206] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.
[0207] Specific embodiments are applied in the present invention to elaborate on the principles and implementation manners of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present invention.
[0208] Those of ordinary skill in the art will realize that the embodiments described herein are provided to assist the reader in understanding the principles of the present invention, and it should be understood that the scope of protection of the present invention is not limited to such specific statements and embodiments. Those of ordinary skill in the art can make various other specific deformations and combinations that do not depart from the essence of the present invention based on these technical revelations disclosed in the present invention, and these deformations and combinations are still within the scope of protection of the present invention.
Claims
1. A method for simulating the state of unstable combustion and overheating of the heating surface of a boiler, characterized in that: The steps include: Step (1): Screening of boiler sensitive parameters: boiler sensitive parameters include parameters related to boiler heating surface overtemperature and parameters related to boiler combustion instability, wherein the parameters related to boiler heating surface overtemperature include discrete operation control parameters and continuous operation control parameters; Step (2): Boiler data collection: Boiler data collection includes boiler combustion instability fault data collection, boiler over-temperature fault data collection and boiler normal operation data collection; Step (3): Boiler data processing: Based on the historical operation data of the unit, the collected operation parameters are processed using statistical and machine learning methods; Step (4): Establishing a state simulation function: the state simulation function integrates a state function, a probability function and a prediction function; Step (5): Evaluate the implementation effect.
2. A method for simulating the state of unstable combustion and overheating of the heating surface of a boiler according to claim 1, characterized in that: The discrete operation control parameters in step (1) are: ① Smoke damper regulating valve position: used to control the smoke flow path; ② Secondary air regulating valve position: used to adjust the amount of secondary air entering the combustion chamber; ③ Gas air regulating valve position: used to adjust the gas air volume entering the burner; The continuous operation control parameters in step (1) are: ①Furnace air volume: affects the oxygen supply in the combustion chamber; ② Economizer inlet flow: affects the water flow entering the economizer and affects the heat exchange efficiency; ③ The steam temperature at the inlet of the first and second stage desuperheaters: affects the cooling effect of the desuperheater; ④ Main steam temperature: directly reflects the temperature of boiler outlet steam; Furthermore, the parameters related to boiler combustion instability specifically include: ①Furnace wind pressure: affects the pressure distribution in the combustion chamber; ② Coal fire detection intensity: reflects the combustion intensity of coal powder in the combustion chamber; ③ Wind speed in the pulverized coal pipeline: affects the distribution and combustion rate of pulverized coal in the combustion chamber; ④ Combustion system operation control parameters: including burner working status, air-coal ratio, fuel input, fuel / air ratio, primary air damper ratio, and secondary air damper opening.
3. The method for simulating the state of unstable combustion and overheating of the heating surface of a boiler according to claim 1, characterized in that: In the step (2), The boiler combustion instability fault data includes the flame detection intensity signals of the 8 burners in each layer of the boiler A-F layers, the coal feed rate of the 6 coal mills, the coal mill current signal and the primary air volume at the inlet of each coal mill, the flow rate of the pulverized coal pipeline in front of each burner and the pulverized coal concentration in the pipeline, the secondary air volume of each layer of the boiler, the burnout air volume, the opening of the secondary air door of each layer, the opening of the tertiary air door of each burner, the opening of the central air door, the opening of the burnout air door, the oxygen content at the tail of the furnace, the total air volume, the total coal volume, the coal quality information and the temperature of the secondary air at the outlet of the air preheater; The boiler over-temperature fault data mainly includes the temperature of each tube wall of the spiral water-cooled wall, the temperature of each tube wall of the vertical water-cooled wall, the temperature of each tube wall of the platen superheater, the temperature of each tube wall of the high-temperature superheater, the temperature of each tube wall of the low-temperature superheater, the temperature of each tube wall of the high-temperature reheater, the temperature of each tube wall of the low-temperature reheater, the high-temperature superheater outlet pressure, boiler load, boiler feed water flow, boiler feed water temperature, total air volume, total fuel volume, main steam parameters, superheater desuperheating water flow, high-temperature superheater outlet temperature, platen superheater inlet / outlet temperature, and the temperature of the air-powder mixture at the pulverizer outlet; The normal operation data of the boiler includes the historical data of boiler load, variable load rate, steam temperature, pressure, air supply, coal supply, coal quality, exhaust gas temperature, oxygen content, fly ash / slag carbon content, burner operation mode, primary air and secondary air parameters.
4. The method for simulating the state of unstable combustion and overheating of the heating surface of a boiler according to claim 1, characterized in that: The processing in step (3) specifically includes the following contents: ① Operating condition identification: By analyzing historical data, the operating parameter characteristics under different operating conditions are identified. The operating data of the boiler under different loads are different. As the load increases, the operating parameters obviously change linearly. ② Abnormal data diagnosis: detect and eliminate abnormal values in the data. Using the standard deviation method, data exceeding ±3 times the standard deviation of the data mean is considered abnormal data; ③ Missing data filling: For missing data, interpolation or prediction methods are used to fill them; ④Data cleaning: remove irrelevant or redundant data.
5. A method for simulating the state of unstable combustion and overheating of the heating surface of a boiler according to claim 4, characterized in that: There are errors in the historical data of the power plant that need to be processed. The errors that need to be processed include systematic errors and random errors. The systematic errors are caused by reading errors, measurement method errors or measurement instrument errors; the random errors are fluctuations in measurement results caused by external interference.
6. A method for simulating the state of unstable combustion and overheating of the heating surface of a boiler according to claim 5, characterized in that: Data preprocessing is performed for system errors. The data preprocessing method is as follows: Assume that the sample data is y1, y2, ..., y n , the average value is The deviation is The standard deviation σ is calculated according to the Bessel formula: When a sample data deviation v k (1≤k≤n) satisfies |v k |>3σ, then it is considered that the data y k It is unreasonable and should be eliminated, where n is the number of data, i represents a certain data, and i is variable; Boiler historical data, the preprocessed data input matrix uses row vector samples. The number of columns in the input matrix represents the total number of model inputs, while the row vector represents each collected data sample.
7. A method for simulating the state of unstable combustion and overheating of the heating surface of a boiler according to claim 5, characterized in that: In order to eliminate the random errors in the measurement data, a smoothing algorithm is used as follows: Assume that some sample data is: {x1,x2,...,x n },x i =s i +n i 4 Where x represents sample data, S i is the true value, n i is the noise. In the above formula, n is the number of data, i represents a certain data, i is variable, then In the formula {a r } is the weight, satisfying y i is the processed data, q is the average value of a distance, is the step size of the window, r represents the value within the distance from -q to q, a r Represents the corresponding weight within the r distance, x i+r Considering the data after the distance r changes, s i+r is a true value after considering the distance r, n i+r is a certain error value after considering the distance r.
8. The method for simulating the state of unstable combustion and overheating of the heating surface of a boiler according to claim 1, characterized in that: In the step (4): ① State function: a state function containing all relevant measuring point data, used to describe the current state of the system, including the selected boiler sensitive parameters; ② Probability function: the probability function of the influence of the measuring point data on the fault, which is used to evaluate the influence of various parameters on the overheating of the heating surface and the instability of combustion; ③ Prediction function: A prediction function for the changing trend of the measurement point data after executing a certain operation, which is used to predict the impact of different operations on the system status.
9. A method for simulating the state of unstable combustion and overheating of the heating surface of a boiler according to claim 8, characterized in that: The specific form of the state simulation function is as follows In the formula, r t is the target variable at time t; s t is the current system status; a t is a discrete decision variable, which is selected from a finite decision set A through a policy function π(·), and t0 represents the initial moment; R(·) is the objective function at time t, reflecting the combustion safety; T(·) is the state transition function, which reflects the change relationship of the system state; C(·) is the state trajectory inequality constraint, where k is an indefinite value ranging from 1 to m; m represents the number of digits; φ(·) is the initial value equality constraint, where l is an indefinite value ranging from 1 to m; m represents the number of digits; h(·) is an action equation, where i is an indefinite value ranging from 1 to p; p represents the number of digits; i represents any data in a set of data; g(·) is an inequality constraint function, where j is an indefinite value ranging from 1 to q; q represents the number of digits.
10. The method for simulating the state of unstable combustion and overheating of the heating surface of a boiler according to claim 1, characterized in that: In step (5), the specific contents of evaluating the implementation effect are: For the input parameters: temperature of each tube wall of spiral water wall, temperature of each tube wall of vertical water wall, temperature of each tube wall of platen superheater, temperature of each tube wall of high temperature superheater, temperature of each tube wall of low temperature superheater, temperature of each tube wall of high temperature reheater, temperature of each tube wall of low temperature reheater, high temperature superheater outlet temperature, boiler load, boiler feed water flow, the output parameter is the predicted temperature value, and the temperature control accuracy is evaluated by comparing the difference between the predicted value and the actual value; According to the input parameters: the flame detection intensity signals of the 8 burners in each layer of the boiler A to F layers, a statistical model based on the flame detection intensity signal is established to calculate the mean and variance of the flame detection intensity signal. If the mean of the flame detection intensity signal decreases or the variance increases, it indicates that the flame detection intensity is unstable or decreasing. A logistic regression model or support vector machine model based on tube wall temperature is established to judge whether the heating surface is over-temperature. The input parameters are the temperature of each tube wall of the spiral water-cooled wall, the temperature of each tube wall of the vertical water-cooled wall, the temperature of each tube wall of the platen superheater, the temperature of each tube wall of the high-temperature superheater, the temperature of each tube wall of the low-temperature superheater, the temperature of each tube wall of the high-temperature reheater, and the temperature of each tube wall of the low-temperature reheater. The output is the over-temperature judgment result. Establish a multivariate linear regression model or a neural network model, with the input parameters being boiler load, variable load rate, steam temperature, pressure, air supply, coal supply, coal quality, exhaust gas temperature, oxygen content, fly ash / slag carbon content, and burner operation mode, and the output parameters being the evaluation index of thermal efficiency or combustion stability; A multivariate linear regression model or a neural network model based on wind parameters is established, and the output is the uniformity evaluation index.
Citation Information
Patent Citations
Method for monitoring thermal state and predicting corrosion of high-temperature heating surface of garbage incinerator
CN117473887A
