A Method and System for Condition Monitoring and Health Assessment Based on Boiler Digital Twin Model

Abstract

This invention discloses a method and system for condition monitoring and health assessment based on a boiler digital twin model. The method includes the following steps: acquiring measurement data of the boiler equipment of a thermal power unit to be monitored and assessed for health; wherein the measurement data includes the steam drum water level, steam flow rate at the steam drum outlet, and steam drum pressure; generating prediction data using a pre-acquired boiler digital twin model of the thermal power unit boiler equipment; wherein the prediction data includes the working fluid mass in the liquid phase zone, steam density, steam drum water enthalpy, steam content in the riser pipe, and steam drum pressure; based on the measurement data and prediction data, performing data assimilation using a volumetric Kalman filter to obtain estimation results; and realizing condition monitoring and health assessment of the boiler equipment of the thermal power unit based on the estimation results. The method provided by this invention can realize online prediction of boiler slagging characteristics.

Description

Technical Field

[0001] This invention belongs to the field of industrial automation technology, specifically the field of boiler equipment monitoring and evaluation, and particularly relates to a method and system for condition monitoring and health assessment based on a boiler digital twin model. Background Technology

[0002] Slagging in thermal power boilers mainly refers to the process where hot ash and unburned pulverized coal in the furnace adhere to the radiant heating surfaces such as water-cooled walls and superheaters in a liquid or semi-liquid state, forming a dense ash layer. This ash has low thermal conductivity and high thermal resistance, which not only reduces heat absorption and transfer within the furnace, causing a decrease in the natural water circulation velocity and an increase in the circulation ratio in the evaporation system, severely impacting boiler output and the stable operation of the evaporation system, but also leads to increased temperatures on unslagged metal heating surfaces and in flue gas, resulting in high-temperature corrosion of the tube walls and overheating of the heating surfaces at the furnace outlet, reducing boiler efficiency and disrupting the overall thermal balance within the furnace. Furthermore, in severe cases of slagging, large slag chunks can detach from the upper part of the furnace, potentially extinguishing the flame or damaging the bottom water-cooled walls, forcing boiler shutdown, or even causing serious accidents such as furnace explosions. Therefore, boiler slagging directly affects the safety, economy, and reliability of unit operation, making condition monitoring and health maintenance of this process of significant practical importance and research value.

[0003] Currently, existing traditional methods, such as the fuzzy mathematical model for comprehensive evaluation of furnace slagging, use seven indicators, including coal ash characteristics and furnace structure and operating parameters, as the judgment criteria. Although this method can comprehensively consider factors related to slagging, it still has the following two limitations:

[0004] (1) Since its judgment index is highly correlated with multiple coal quality components such as softening degree, silicon-aluminum ratio, silicon ratio and acid-base ratio, its accuracy and reliability will decrease when the coal quality of the boiler is uneven or mixed coal combustion is used.

[0005] (2) For the parameters of multiple judgment indicators in the model, it is difficult to obtain accurate values ​​in real time in engineering. Most of them rely on past experience to make calculations, which increases the prediction error of the model. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for condition monitoring and health assessment based on a boiler digital twin model, in order to solve one or more of the aforementioned technical problems. Specifically, the method provided by this invention is an online condition health prediction method for boiler equipment in thermal power units based on digital twins, which can realize online prediction of boiler slagging characteristics.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] This invention provides a method for condition monitoring and health assessment based on a boiler digital twin model, comprising the following steps:

[0009] Acquire measurement data of boiler equipment in thermal power units that are subject to condition monitoring and health assessment; wherein, the measurement data includes steam drum water level, steam drum outlet steam flow rate, and steam drum pressure;

[0010] Predictive data is generated using a pre-acquired digital twin model of the boiler equipment of the thermal power unit; wherein, the predicted data includes the working fluid mass in the liquid phase region, steam density, enthalpy of water in the steam drum, steam content in the riser pipe, and steam drum pressure;

[0011] Based on the measurement data and the prediction data, data assimilation is performed using capacitive Kalman filtering to obtain the estimation results;

[0012] Based on the estimation results, condition monitoring and health assessment of boiler equipment in thermal power units can be achieved.

[0013] A further improvement to the method of the present invention is that the boiler digital twin model of the thermal power unit boiler equipment is represented as follows:

[0014]

[0015] In the formula, x k t k u k and z k+1 These are the system state vector, parameter vector, input vector, and measurement vector, respectively; A, B, and C are all coefficient matrices; ω k and ξ k All are process noise, v k To measure noise, ω k ~N(0,Q) k ), ξ k ~N(0,G k ), v k ~N(0,R k+1 );Q k and G k All are process noise covariance matrices, R k+1 The measurement noise covariance matrix;

[0016] x k t k u k and z k+1 They are expressed as follows:

[0017] x k =[M dl ρ v h w q v Pdr ] T

[0018] t k =[K sl ]

[0019] u k =[W e T b P s ] T

[0020] z k+1 =[LW v P dr ] T

[0021] The coefficient matrices A, B, and C are represented as follows:

[0022]

[0023]

[0024]

[0025] In the formula,

[0026] a 11 =1,

[0027]

[0028]

[0029]

[0030]

[0031]

[0032]

[0033]

[0034]

[0035]

[0036]

[0037]

[0038]

[0039]

[0040]

[0041]

[0042]

[0043]

[0044]

[0045]

[0046] M dl For the working fluid in the liquid phase region, ρ v h is the saturated vapor density. w q represents the enthalpy of the water in the steam drum. v P represents the mass vapor content of a carbonated water mixture. dr For steam drum pressure; K sl The heat transfer coefficient of a water-cooled wall, W e For water supply flow rate, T b P represents the average temperature of the furnace. s Where L is the superheater pressure, W is the steam drum water level, and W is the water level in the drum. v T is the steam flow rate at the steam drum outlet. s Sampling time; W ro W is the flow rate of the steam-water mixture at the riser outlet. d For the downcomer inlet flow rate, W ec V represents the dynamic evaporation rate in the liquid phase region. v V is the volume of the vapor phase region. L ρ is the volume of the liquid phase region. w h is the density of water inside the steam drum. e h is the enthalpy value of the feedwater. wv h represents the enthalpy of saturated water at the riser outlet. v K represents the enthalpy of saturated vapor. ec T is an empirical coefficient. w T represents the temperature of the steam drum water. v V is the saturated steam temperature. r T is the volume of the riser pipe. r W represents the average temperature of the riser pipe. d To reduce the flow rate in the downpipe, W ro ρ is the flow rate at the riser outlet. r R is the density of the carbonated beverage mixture. f Q represents the resistance to flow in the pipe. sl A represents the total heat absorbed by the riser pipe. L B L C L DL A dr B dr C dr D dr A ve B ve C ve D ve A vt B vt C vt D vt E vt A wv B wv C wv D wv E wv A wt B wt C wt D wt A wd B wd C wd and D wd All are fitting coefficients.

[0047] A further improvement to the method of the present invention is that the step of assimilating the data using capacitive Kalman filtering based on the measurement data and the prediction data to obtain the estimation result includes:

[0048] First, the filtering algorithm is initialized by processing the state variables, parameter values, noise covariance matrix, and model coefficient matrices x0, t0, u0, P0, and Q. k G k R k+1 Set initial values ​​for A, B, and C;

[0049] Secondly, the calculation steps for the time update process are as follows:

[0050] Calculate the square root of variance S k and volume point [X j,k T j,k ] T The expressions are as follows:

[0051] S k =chol{P k}

[0052]

[0053] In the formula, chole represents the Cholesky decomposition of the matrix, n is the dimension of the state variables, and P k For the filter covariance, and The state and the filtered parameter value at time k are respectively, ξ jFor the basic volume point, the expression is:

[0054]

[0055] In the formula, [I] represents the set of points generated by permuting and changing the signs of the elements of a unit vector, which is a complete and fully symmetric set of points; [I] j Represents the j-th point in the point set;

[0056] Calculate the volume point propagated through the nonlinear state-space model. The expression is:

[0057]

[0058] In the formula, f(·) is a nonlinear state-space model;

[0059] One-step prediction value based on propagation volume point calculation state and parameters. And one-step prediction of covariance P k+1|k The expression is:

[0060]

[0061] In the formula, ω j The weights corresponding to the volume points are expressed as follows:

[0062] Finally, the calculation steps for the measurement update process are as follows:

[0063] Calculate the square root of the prediction variance S based on the one-step prediction value and the one-step prediction covariance. k+1|k With the predicted volume point [X] j,k+1 T j,k+1 ] T The expression is:

[0064] S k+1|k =chol{P k+1|k}

[0065]

[0066] Calculate the volume point Z propagated through the nonlinear measurement equation model. j,k+1 The expression is:

[0067]

[0068] Calculate the predicted value of the measurement Measurement error covariance P zz,k+1 With mutual covariance P xz,k+1 The expression is:

[0069]

[0070] Calculate the Kalman gain K k+1 State and parameter estimates and covariance P k+1 The expression is:

[0071]

[0072] In the formula, and The estimated results for the state and parameters are used for condition monitoring and health assessment of the boiler system.

[0073] A further improvement to the method of the present invention is that the step of realizing condition monitoring and health assessment of boiler equipment of thermal power units based on the estimation results specifically includes:

[0074] The degree of degradation of the heat transfer coefficient of the water-cooled wall ζ sl Represented as:

[0075] In the formula, K represents the heat transfer coefficient of the water-cooled wall under rated load when there is no slag buildup inside the furnace. sl The actual heat transfer coefficient of the water-cooled wall under rated load represents the evaluation result;

[0076] Based on the operating conditions, an evaluation standard for in-furnace slagging characteristics based on the degree of heat transfer coefficient degradation is obtained, expressed as:

[0077]

[0078] In the formula, S l and S h The threshold value is pre-calibrated.

[0079] This invention provides a condition monitoring and health assessment system based on a boiler digital twin model, comprising:

[0080] The measurement data acquisition module is used to acquire measurement data of boiler equipment in thermal power units that are to be monitored for status and assessed for health; wherein, the measurement data includes steam drum water level, steam drum outlet steam flow rate and steam drum pressure;

[0081] The prediction data acquisition module is used to generate prediction data using a pre-acquired digital twin model of the boiler equipment of the thermal power unit; wherein, the prediction data includes the working fluid mass in the liquid phase region, steam density, enthalpy of water in the steam drum, steam content in the riser pipe, and steam drum pressure;

[0082] The evaluation result acquisition module is used to assimilate the data using a capacitive Kalman filter based on the measurement data and the prediction data to obtain the estimation result.

[0083] The monitoring and evaluation module is used to monitor the status and assess the health of boiler equipment in thermal power units based on the estimation results.

[0084] Compared with the prior art, the present invention has the following beneficial effects:

[0085] To address the technical problems of existing methods, this invention provides a digital twin online state health prediction method for boiler equipment in thermal power units. First, a digital twin model of the boiler evaporation system is established through mechanistic analysis. Second, combined with measurement data from the physical boiler equipment, a real-time data assimilation method based on volumetric Kalman filtering is constructed to synchronously reflect the dynamic changes of the physical equipment online. Finally, a real-time estimation model of the water-cooled wall heat transfer coefficient is established using key state parameters of the twin system, such as steam drum pressure, steam drum outlet evaporation rate, riser average temperature, and furnace average temperature. By analyzing the degree of heat transfer coefficient degradation, online prediction of boiler slagging characteristics is achieved. Exemplary results show that the method of this invention can accurately reflect the change process of the heat transfer coefficient and effectively determine the slagging level in the furnace based on the degree of degradation of the water-cooled wall heat transfer coefficient, thereby improving boiler operating efficiency and ensuring the safe operation of the unit. Attached Figure Description

[0086] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art are briefly introduced below; obviously, the drawings described below are some embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings without creative effort.

[0087] Figure 1 This is a flowchart illustrating a method for state monitoring and health assessment based on a boiler digital twin model according to an embodiment of the present invention.

[0088] Figure 2 This is a schematic diagram of the data assimilation and parameter estimation process of a digital twin system in an embodiment of the present invention;

[0089] Figure 3 This is a schematic diagram showing the comparison between the actual and estimated values ​​of the heat transfer coefficient of the water-cooled wall in an embodiment of the present invention.

[0090] Figure 4 This is a schematic diagram comparing the actual and estimated values ​​of the vapor content of the riser pipe in an embodiment of the present invention.

[0091] Figure 5 This is a schematic diagram showing the comparison between the actual and estimated values ​​of the steam drum pressure in an embodiment of the present invention.

[0092] Figure 6This is a schematic diagram showing the comparison between the actual and estimated values ​​of saturated vapor density in an embodiment of the present invention;

[0093] Figure 7 This is a schematic diagram comparing the actual and estimated values ​​of the mass in the liquid phase region of the steam drum in an embodiment of the present invention.

[0094] Figure 8 This is a schematic diagram comparing the actual and estimated values ​​of the enthalpy of the water in the steam drum in an embodiment of the present invention;

[0095] Figure 9 This is a curve comparing the actual and estimated values ​​of the heat transfer coefficient degradation degree and a schematic diagram of the slagging level prediction in an embodiment of the present invention. Detailed Implementation

[0096] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0097] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0098] The present invention will now be described in further detail with reference to the accompanying drawings:

[0099] An embodiment of the present invention provides a method for condition monitoring and health assessment based on a boiler digital twin model, which specifically includes the following steps:

[0100] Acquire measurement data of boiler equipment in thermal power units that are subject to condition monitoring and health assessment; the measurement data includes steam drum water level, steam drum outlet steam flow rate, and steam drum pressure.

[0101] Predictive data is generated using a pre-acquired digital twin model of the boiler equipment of the thermal power unit; the predicted data includes the working fluid mass in the liquid phase region, steam density, enthalpy of water in the steam drum, steam content in the riser pipe, and steam drum pressure;

[0102] Based on the measurement data and the prediction data, data assimilation is performed using capacitive Kalman filtering to obtain the estimation results;

[0103] Based on the estimation results, condition monitoring and health assessment of boiler equipment in thermal power units can be achieved.

[0104] Specifically, as a foundation for subsequent work, this invention first establishes a digital twin mechanism model of the boiler system and a calculation equation for the heat transfer coefficient of the water-cooled wall, synchronously describing the dynamic changes of the physical equipment in real time. For illustrative purposes, the boiler evaporation system mainly includes a steam drum, downcomer, and riser. Feedwater from the economizer enters the steam drum and flows into the downcomer, then through the header into the riser to absorb heat from the metal wall and heat the feedwater, converting some of the water into saturated steam, i.e., a steam-water mixture. This mixture rises to the steam drum for steam-water separation. The saturated steam flows out to the superheater, while the saturated water re-enters the riser through the downcomer for reheating, forming a natural circulation.

[0105] Assuming the working fluid inside the steam drum is completely saturated, the mass balance equations for the liquid and vapor phases of the steam drum are expressed as follows:

[0106]

[0107]

[0108] In the formula, W e For water supply flow rate, W ro W is the flow rate of the steam-water mixture at the riser outlet. d For the downcomer inlet flow rate, W ec q represents the dynamic evaporation rate in the liquid phase region. v M represents the mass vapor content of a carbonated water mixture. dl For the working fluid mass in the liquid phase region, W v V is the steam flow rate at the steam drum outlet. v Let ρ be the volume of the vapor phase region. v This is the density of saturated vapor.

[0109] The water level in the steam drum is related to the mass and density of the working fluid in the liquid phase region and the shape of the steam drum, and can be expressed as:

[0110]

[0111] In the formula, V L ρ is the volume of the liquid phase region, L is the water level in the steam drum, and ρ is the volume of the liquid phase region. w This is the density of water inside the steam drum.

[0112] V L The functional relationship between L and the boiler drum depends on the shape of the boiler drum. In practice, the boiler drum is generally a horizontally placed cylinder with hemispherical heads at both ends, so it can be written as:

[0113]

[0114] In the formula, r is the radius of the spherical head, H is the total length of the steam drum, and θ is the angle between the water level and the center of the circle;

[0115] The total volume of the steam drum is:

[0116]

[0117] By solving the inverse function L = f of equation (6) -1 (V L Calculating the water level in the steam drum is too complex and not conducive to model solving. Therefore, this invention uses a function fitting method to obtain an approximate water level-liquid phase volume function relationship, expressed as:

[0118]

[0119] Among them, A L B L C L and D L All are fitting coefficients.

[0120] The energy balance equation for the liquid phase region of the steam drum is:

[0121]

[0122] Among them, h e h is the enthalpy value of the feedwater. wv h represents the enthalpy of saturated water at the riser outlet. w h is the enthalpy of water in the steam drum. v This is the enthalpy of saturated vapor.

[0123] Dynamic evaporation rate W ec Represented as:

[0124] W ec =K ec ·(T w -T v (9)

[0125] Among them, K ec T is an empirical coefficient. w T represents the temperature of the steam drum water. v This is the saturated steam temperature.

[0126] The steam drum water enters the riser from the downcomer through the header, where it absorbs radiant heat from the furnace and evaporates into a steam-water mixture. Its mass balance equation is expressed as:

[0127]

[0128] In the formula, V r For the riser volume, W d To reduce the flow rate in the downpipe, W ro ρ is the flow rate at the riser outlet. r For the density of the soda-water mixture, this invention simplifies the calculation by using the average temperature T of the riser pipe. r Represented as:

[0129]

[0130] Referring to relevant books on boiler principles, the steam content q of the riser pipe mass is... v Steam flow rate W at the steam drum outlet v There exists a linear functional relationship between them, written as:

[0131]

[0132] Among them, A q and B q The coefficients to be identified This refers to the steam content per unit mass corresponding to the rated load condition. W represents the steam flow rate at the steam drum outlet under rated load conditions. v It can be obtained from the following formula:

[0133]

[0134] Among them, P dr For the steam drum pressure, P s R is the superheater pressure. f The resistance to flow in the pipe. Under steady-state conditions, the radiant heat from the furnace in the riser pipe can be divided into two parts: one part is heat Q. rs Used to heat the inlet water of the riser pipe to saturation water, another portion of the heat Q zf Used to heat saturated water into saturated steam, that is:

[0135] Q rs =W d ·(h wv -h w (14)

[0136] Q zf =q v ·W ro ·(h v -h wv (15)

[0137] Based on the radiative heat transfer formula, we can calculate an estimation model for the heat transfer coefficient of the water-cooled wall:

[0138]

[0139] Among them, T b Q represents the average temperature of the furnace. sl The total heat absorbed by the riser pipe is:

[0140] Q sl =Q rs +Q zf (17)

[0141] In the simulation of this invention, to simplify the calculation, it is generally assumed that all parameters in the evaporation system are working fluid parameters in a saturated state. The thermodynamic property tables of saturated water and saturated steam can be referenced.

[0142] According to the thermodynamic property table, the saturated vapor density ρ in this embodiment of the invention... v Calculate the steam drum pressure P dr enthalpy of steam h v Steam temperature T v and the enthalpy of saturated water h wv The enthalpy h of the steam drum water w Calculate the temperature T of the steam drum water w And the density ρ of water in the steam drum w The polynomials of each function are expressed as follows:

[0143]

[0144]

[0145]

[0146]

[0147]

[0148]

[0149] Based on the aforementioned boiler system mechanism equations, a nonlinear discrete state-space model with parameter estimation is constructed, written as:

[0150]

[0151] In the formula, x k t k u k and z k+1 These are the system state vector, parameter vector, input vector, and measurement vector, respectively; A, B, and C are coefficient matrices; ω kand ξ k For process noise, v k To measure noise, the three factors are independent of each other, and ω k ~N(0,Q) k ), ξ k ~N(0,G k ), v k ~N(0,R k+1 );

[0152] x k t k u k and z k+1 They are expressed as follows:

[0153] x k =[M dl ρ v h w q v P dr ] T (25)

[0154] t k =[K sl (26)

[0155] u k =[W e T b P s ] T (27)

[0156] z k+1 =[LW v P dr ] T (28)

[0157] The coefficient matrices A, B, and C are written as follows:

[0158]

[0159]

[0160]

[0161] In the formula,

[0162]

[0163]

[0164]

[0165]

[0166]

[0167]

[0168]

[0169] In the formula, Equation (7) represents the functional relationship between the steam drum water level and the volume of the liquid phase region. and Represent the functional relationships of equations (18) to (23) respectively, T s Sampling time.

[0170] Please see Figure 2 In this embodiment of the invention, real-time data assimilation and online estimation of key parameters of a digital twin system are achieved based on the capacitive Kalman filter (CKF) algorithm; the detailed calculation process of the CKF algorithm can be found in relevant literature (exemplary, Cubature Kalman Filters).

[0171] When using the capacitive Kalman filter algorithm, the specific calculation process includes:

[0172] First, the filtering algorithm is initialized by processing the state variables, parameter values, noise covariance matrix, and model coefficient matrices x0, t0, u0, P0, and Q. k G k R k+1 Set initial values ​​for A, B, and C;

[0173] Secondly, the calculation steps for the time update process are as follows:

[0174] 1) Calculate the square root of variance S k and volume point [X j,k T j,k ] T

[0175] S k =chol{P k} (32)

[0176]

[0177] In the formula, chole represents the Cholesky decomposition of the matrix, n is the dimension of the state variables, and P k For the filter covariance, and The state and the filtered parameter value at time k are respectively, ξ j The basic volume point, i.e.:

[0178]

[0179] In the formula, [I] represents the set of points generated by permuting and changing the signs of the elements of a unit vector, called the complete fully symmetric set of points. j It represents the j-th point in the set of points.

[0180] 2) Calculate the volume point propagated through the nonlinear state-space model.

[0181]

[0182] In the formula, f(·) is the nonlinear state-space model described in formula (24).

[0183] 3) Calculate the state and parameter one-step prediction values ​​based on the propagation volume point. And one-step prediction of covariance P k+1|k

[0184]

[0185] In the formula, ω j The weights corresponding to the volume points:

[0186] Finally, the calculation steps for the measurement update process are as follows:

[0187] 1) Calculate the root of the predicted variance S based on the one-step predicted value and the one-step predicted covariance. k+1|k With the predicted volume point [X] j,k+1 T j,k+1 ] T

[0188] S k+1|k =chol{P k+1|k} (38)

[0189]

[0190] 2) Calculate the volume point Z propagated through the nonlinear measurement equation model. j,k+1

[0191]

[0192] 3) Calculate the predicted measurement value Measurement error covariance P zz,k+1 With mutual covariance P xz,k+1

[0193]

[0194] 4) Calculate the Kalman gain K k+1 State and parameter estimates and covariance P k+1

[0195]

[0196] In the formula, and The results are the estimation results of the state and parameters, respectively, and are used for the condition monitoring and health assessment of the boiler system.

[0197] In this embodiment of the invention, the steps for realizing the status monitoring and health assessment of boiler equipment in thermal power units based on the estimation results may specifically include: establishing a prediction model for slagging characteristics in the furnace to realize real-time status monitoring and online health assessment of the boiler system.

[0198] Because ash has a low thermal conductivity and high thermal resistance, it reduces heat absorption and deteriorates heat transfer within the furnace. Specifically, as the ash layer thickens, the heat transfer coefficient of the water-cooled wall gradually decreases; in other words, the degradation of the water-cooled wall's heat transfer performance is one of the most significant characteristics of slagging within the furnace. Therefore, unlike existing methods, this invention predicts the level of slagging within the furnace based on the degree of degradation of the water-cooled wall's heat transfer coefficient. The degree of degradation of the water-cooled wall's heat transfer coefficient is ζ. sl It can be written as:

[0199]

[0200] In the formula, K represents the heat transfer coefficient of the water-cooled wall under rated load when there is no slag buildup inside the furnace. sl This refers to the heat transfer coefficient of the water-cooled wall during actual boiler operation. Based on the actual boiler operation, an evaluation standard for in-furnace slagging characteristics based on the degree of heat transfer coefficient degradation can be obtained:

[0201]

[0202] In the formula, S l and S h The specific values ​​can be calibrated according to the actual boiler.

[0203] The following are embodiments of the apparatus of the present invention, which can be used to execute embodiments of the method of the present invention. For details not omitted in the apparatus embodiments, please refer to the embodiments of the method of the present invention.

[0204] This invention provides a condition monitoring and health assessment system based on a boiler digital twin model, comprising:

[0205] The measurement data acquisition module is used to acquire measurement data of boiler equipment in thermal power units that are to be monitored for status and assessed for health; wherein, the measurement data includes steam drum water level, steam drum outlet steam flow rate and steam drum pressure;

[0206] The prediction data acquisition module is used to generate prediction data using a pre-acquired digital twin model of the boiler equipment of the thermal power unit; wherein, the prediction data includes the working fluid mass in the liquid phase region, steam density, enthalpy of water in the steam drum, steam content in the riser pipe, and steam drum pressure;

[0207] The evaluation result acquisition module is used to assimilate the data using a capacitive Kalman filter based on the measurement data and the prediction data to obtain the estimation result.

[0208] The monitoring and evaluation module is used to monitor the status and assess the health of boiler equipment in thermal power units based on the estimation results.

[0209] The system provided in this invention establishes a digital twin model of the boiler combustion and evaporation system through mechanistic analysis. Combined with measurement data from the physical boiler equipment, a real-time data assimilation method based on volumetric Kalman filtering is constructed to synchronously reflect the dynamic changes of the physical equipment online. A real-time estimation model of the water-cooled wall heat transfer coefficient is established using key state variables such as the drum pressure, drum outlet steam flow rate, riser average temperature, and furnace average temperature of the twin system, enabling state monitoring and health assessment of the boiler system. This invention's digital twin data assimilation method for boiler equipment based on volumetric Kalman filtering achieves online correction and real-time estimation of key state variables and parameters, effectively suppressing system noise and completing the state monitoring and health assessment of the physical equipment.

[0210] In a specific embodiment of the present invention, in conjunction with the technical solutions disclosed in the above embodiments, the effectiveness of the model and method is verified by analyzing experimental data results; wherein, the initial value of the state variable x0, the initial value of the parameter t0, the initial value of the input variable u0, the initial value of the state variable covariance matrix P0, and the initial value of the process noise covariance matrix Q are... k and G k Measurement noise covariance matrix R k+1 Selected respectively as:

[0211] x0=[38145.71 146.85 1593.44 0.2079 18.83] T (45)

[0212] t0 = [4.80 × 10 -7 (46)

[0213] u0 = [538.17 1086.49 18.15] T (47)

[0214] G k =1×10 -22 (48)

[0215]

[0216]

[0217]

[0218] In this embodiment of the invention, please refer to Table 1 for the selection of corresponding parameter values ​​in the system model.

[0219] Table 1. Selection of Model Parameter Values

[0220]

[0221]

[0222] Please see Figures 3 to 8 This invention simulates the slow decrease in the heat transfer coefficient of a water-cooled wall and observes the boiler system state variable M. dl ρ v h w q v P dr and parameter K sl The dynamic change process and real-time estimation results. In actual boiler operation, the heat transfer coefficient K of the water-cooled wall is affected by slagging within the furnace. sl A slow decline, the actual process of change is as follows Figure 3 As shown by the solid line, the dashed line represents the estimation result of the CKF algorithm. It can be seen that the estimation result can effectively track changes in the actual value, accurately reflecting the change process of the heat transfer coefficient in real time. Due to the decrease in the heat transfer coefficient, the total heat absorption in the riser pipe decreases, leading to a decrease in the mass vapor content q of the steam-water mixture in the riser pipe. v Decrease, such as Figure 4 As shown, a decrease in the mass steam content leads to a decrease in steam flow rate, resulting in a decrease in the steam drum pressure P. dr Decrease, such as Figure 5 As shown in the table of thermodynamic properties of saturated water and saturated steam, as the steam drum pressure decreases, the steam density ρ... v It will decrease accordingly, such as Figure 6 As shown, a decrease in steam density leads to a decrease in saturated steam temperature, increasing the dynamic evaporation rate of water in the steam drum and causing a decrease in the mass M of the liquid phase region of the steam drum. dl Decrease, such as Figure 7 As shown, simultaneously, an increase in the proportion of saturated water flow at the riser outlet will lead to an increase in the enthalpy h of the steam drum water. w Ascending, such as Figure 8 As shown in Table 2, the estimated values ​​of all variables can quickly and accurately track changes in actual values, while effectively suppressing system noise and measurement noise, demonstrating excellent filtering and estimation performance. The performance indicators for state and parameter estimation are shown in Table 2. Therefore, the data assimilation method used in this invention can meet the requirements for state monitoring and health assessment of digital twin systems.

[0223] Table 2. Performance Indicators of State and Parameter Estimation

[0224]

[0225] Please see Figure 9 , Figure 9 The degree of heat transfer coefficient degradation ζ sl The prediction results show that the two thick solid lines represent the slagging grade dividing lines in formula (44), and are obtained by ζ. sl The current slagging level of the boiler can be quickly determined within the specified range, facilitating timely maintenance and cleaning of related equipment by staff, improving boiler efficiency, ensuring safe operation of the unit, and reducing maintenance costs.

[0226] In summary, existing methods use coal ash composition analysis, including multiple parameters such as softening degree, silica-alumina ratio, silica ratio, and acid-base ratio, as important indicators for determining slagging characteristics. However, for boilers burning uneven coal quality or using mixed coal combustion, the accuracy and reliability of predictions decrease, and corresponding testing and analysis instruments are required, increasing prediction costs. To address this issue, this invention establishes a digital twin dynamic mechanism model of the boiler system, constructs an estimation model of the water-cooled wall heat transfer coefficient based on key state variables, and predicts the slagging level in the furnace based on the degree of heat transfer coefficient degradation. This effectively avoids the ash composition analysis process and improves the prediction accuracy of slagging characteristics under uneven coal quality or mixed coal combustion conditions. Specifically, this invention establishes a soft measurement method for the water-cooled wall heat transfer coefficient based on easily obtainable state variables such as steam drum outlet evaporation rate and steam drum pressure, avoiding the measurement process of complex and variable parameters and the purchase of testing instruments, thus reducing the prediction cost of in-furnace slagging characteristics.

[0227] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0228] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0229] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0230] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0231] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for condition monitoring and health assessment based on a boiler digital twin model, characterized in that, Includes the following steps: Acquire measurement data of boiler equipment in thermal power units that are subject to condition monitoring and health assessment; wherein, the measurement data includes steam drum water level, steam drum outlet steam flow rate, and steam drum pressure; Predictive data is generated using a pre-acquired digital twin model of the boiler equipment of the thermal power unit; wherein, the predicted data includes the working fluid mass in the liquid phase region, steam density, enthalpy of water in the steam drum, steam content in the riser pipe, and steam drum pressure; Based on the measurement data and the prediction data, data assimilation is performed using capacitive Kalman filtering to obtain the estimation results; Based on the estimation results, condition monitoring and health assessment of boiler equipment in thermal power units can be achieved. in, The boiler digital twin model of the thermal power unit boiler equipment is represented as follows: In the formula, x k t k u k and z k+1 These are the system state vector, parameter vector, input vector, and measurement vector, respectively; A, B, and C are all coefficient matrices; ω k and ξ k All are process noise, v k+1 For the measurement noise at time k+1, ω k ~N(0,Q) k ), ξ k ~N(0,G k ), v k ~N(0,R k+1 );Q k and G k All are process noise covariance matrices, R k+1 The measurement noise covariance matrix; x k t k u k and z k+1 They are expressed as follows: x k =[M dl ρ v h w q v P dr ] T t k =[K sl ] u k =[W e T b P s ] T z k+1 =[L W v P dr ] T The coefficient matrices A, B, and C are represented as follows: In the formula, a 11 ＝1， M dl For the working fluid in the liquid phase region, ρ v h is the saturated vapor density. w q represents the enthalpy of the water in the steam drum. v P represents the mass vapor content of a carbonated water mixture. dr For steam drum pressure; K sl The heat transfer coefficient of a water-cooled wall, W e For water supply flow rate, T b P represents the average temperature of the furnace. s Where L is the superheater pressure, W is the steam drum water level, and W is the water level in the drum. v T is the steam flow rate at the steam drum outlet. s Sampling time; W ro W is the flow rate of the steam-water mixture at the riser outlet. ec V represents the dynamic evaporation rate in the liquid phase region. v V is the volume of the vapor phase region. L ρ is the volume of the liquid phase region. w h is the density of water inside the steam drum. e h is the enthalpy value of the feedwater. wv h represents the enthalpy of saturated water at the riser outlet. v K represents the enthalpy of saturated vapor. ec T is an empirical coefficient. w T represents the temperature of the steam drum water. v V is the saturated steam temperature. r T is the volume of the riser pipe. r W is the average temperature of the riser pipe. ro ρ is the flow rate at the riser outlet. r R is the density of the carbonated beverage mixture. f Q represents the resistance to flow in the pipe. sl A represents the total heat absorbed by the riser pipe. L B L C L D L A dr B dr C dr D dr A ve B ve C ve D ve A vt B vt C vt D vt E vt A wv B wv C wv D wv E wv A wt B wt C wt D wt A wd B wd C wd and D wd All are fitting coefficients; The step of assimilating data using capacitive Kalman filtering based on the measurement data and the prediction data to obtain the estimation result includes: First, the filtering algorithm is initialized by processing the state variables, parameter values, noise covariance matrix, and model coefficient matrices x0, t0, u0, P0, and Q. k G k R k+1 Set initial values ​​for A, B, and C; Secondly, the calculation steps for the time update process are as follows: Calculate the square root of variance S k and volume point [X j,k T j,k ] T The expressions are as follows: S k =chol{P k } In the formula, chole represents the Cholesky decomposition of the matrix, n is the dimension of the state variables, and P k For the filter covariance, and The state and the filtered parameter value at time k are respectively, ξ j For the basic volume point, the expression is: In the formula, [I] represents the set of points generated by permuting and changing the signs of the elements of a unit vector, which is a complete and fully symmetric set of points; [I] j Represents the j-th point in the point set; Calculate the volume point propagated through the nonlinear state-space model. The expression is: In the formula, f(·) is a nonlinear state-space model; One-step prediction value based on propagation volume point calculation state and parameters. And one-step prediction of covariance P k+1|k The expression is: In the formula, ω j The weights corresponding to the volume points are expressed as follows: Finally, the calculation steps for the measurement update process are as follows: Calculate the square root of the prediction variance S based on the one-step prediction value and the one-step prediction covariance. k+1|k With the predicted volume point [X] j,k+ 1T j,k+1 ] T The expression is: S k+1|k =chol{P k+1|k } Calculate the volume point Z propagated through the nonlinear measurement equation model. j,k+1 The expression is: Calculate the predicted value of the measurement Measurement error covariance P zz,k+1 With mutual covariance P xz,k+1 The expression is: Calculate the Kalman gain K k+1 State and parameter estimates and covariance P k+1 The expression is: In the formula, and The estimated results for the state and parameters are used for condition monitoring and health assessment of the boiler system. The steps for implementing condition monitoring and health assessment of boiler equipment in thermal power units based on the estimation results specifically include: The degree of degradation of the heat transfer coefficient of the water-cooled wall ζ sl Represented as: In the formula, K represents the heat transfer coefficient of the water-cooled wall under rated load when there is no slag buildup inside the furnace. sl The actual heat transfer coefficient of the water-cooled wall under rated load represents the evaluation result; Based on the operating conditions, an evaluation standard for in-furnace slagging characteristics based on the degree of heat transfer coefficient degradation is obtained, expressed as: In the formula, S l and S h The threshold value is pre-calibrated.

2. A condition monitoring and health assessment system based on a boiler digital twin model, characterized in that, include: The measurement data acquisition module is used to acquire measurement data of boiler equipment in thermal power units that are to be monitored for status and assessed for health; wherein, the measurement data includes steam drum water level, steam drum outlet steam flow rate and steam drum pressure; The prediction data acquisition module is used to generate prediction data using a pre-acquired digital twin model of the boiler equipment of the thermal power unit; wherein, the prediction data includes the working fluid mass in the liquid phase region, steam density, enthalpy of water in the steam drum, steam content in the riser pipe, and steam drum pressure; The evaluation result acquisition module is used to assimilate the data using a capacitive Kalman filter based on the measurement data and the prediction data to obtain the estimation result. The monitoring and evaluation module is used to monitor the status and assess the health of boiler equipment in thermal power units based on the estimation results. in, The boiler digital twin model of the thermal power unit boiler equipment is represented as follows: In the formula, x k t k u k and z k+1 These are the system state vector, parameter vector, input vector, and measurement vector, respectively; A, B, and C are all coefficient matrices; ω k and ξ k All are process noise, v k+1 For the measurement noise at time k+1, ω k ~N(0,Q) k ), ξ k ~N(0,G k ), v k ~N(0,R k+1 );Q k and G k All are process noise covariance matrices, R k+1 The measurement noise covariance matrix; x k t k u k and z k+1 They are expressed as follows: x k =[M dl ρ v h w q v P dr ] T t k =[K sl ] u k =[W e T b P s ] T z k+1 =[L W v P dr ] T The coefficient matrices A, B, and C are represented as follows: In the formula, a 11 ＝1， M dl For the working fluid in the liquid phase region, ρ v h is the saturated vapor density. w q represents the enthalpy of the water in the steam drum. v P represents the mass vapor content of a carbonated water mixture. dr For steam drum pressure; K sl The heat transfer coefficient of a water-cooled wall, W e For water supply flow rate, T b P represents the average temperature of the furnace. s Where L is the superheater pressure, W is the steam drum water level, and W is the water level in the drum. v T is the steam flow rate at the steam drum outlet. s Sampling time; W ro W is the flow rate of the steam-water mixture at the riser outlet. ec V represents the dynamic evaporation rate in the liquid phase region. v V is the volume of the vapor phase region. L ρ is the volume of the liquid phase region. w h is the density of water inside the steam drum. e h is the enthalpy value of the feedwater. wv h represents the enthalpy of saturated water at the riser outlet. v K represents the enthalpy of saturated vapor. ec T is an empirical coefficient. w T represents the temperature of the steam drum water. v V is the saturated steam temperature. r T is the volume of the riser pipe. r W represents the average temperature of the riser pipe. ro ρ is the flow rate at the riser outlet. r R is the density of the carbonated beverage mixture. f Q represents the resistance to flow in the pipe. sl A represents the total heat absorbed by the riser pipe. L B L C L D L A dr B dr C dr D dr A ve B ve C ve D ve A vt B vt C vt D vt E vt A wv B wv C wv D wv E wv A wt B wt C wt D wt A wd B wd C wd and D wd All are fitting coefficients; The step of assimilating data using capacitive Kalman filtering based on the measurement data and the prediction data to obtain the estimation result includes: First, the filtering algorithm is initialized by processing the state variables, parameter values, noise covariance matrix, and model coefficient matrices x0, t0, u0, P0, and Q. k G k R k+1 Set initial values ​​for A, B, and C; Secondly, the calculation steps for the time update process are as follows: Calculate the square root of variance S k and volume point [X j,k T j,k ] T The expressions are as follows: S k =chol{P k } In the formula, chole represents the Cholesky decomposition of the matrix, n is the dimension of the state variables, and P k For the filter covariance, and The state and the filtered parameter value at time k are respectively, ξ j For the basic volume point, the expression is: In the formula, [I] represents the set of points generated by permuting and changing the signs of the elements of a unit vector, which is a complete and fully symmetric set of points; [I] j Represents the j-th point in the point set; Calculate the volume point propagated through the nonlinear state-space model. The expression is: In the formula, f(·) is a nonlinear state-space model; One-step prediction value based on propagation volume point calculation state and parameters. And one-step prediction of covariance P k+1|k The expression is: In the formula, ω j The weights corresponding to the volume points are expressed as follows: Finally, the calculation steps for the measurement update process are as follows: Calculate the square root of the prediction variance S based on the one-step prediction value and the one-step prediction covariance. k+1|k With the predicted volume point [X] j,k+ 1T j,k+1 ] T The expression is: S k+1|k =chol{P k+1|k } Calculate the volume point Z propagated through the nonlinear measurement equation model. j,k+1 The expression is: Calculate the predicted value of the measurement Measurement error covariance P zz,k+1 With mutual covariance P xz,k+1 The expression is: Calculate the Kalman gain K k+1 State and parameter estimates and covariance P k+1 The expression is: In the formula, and The estimated results for the state and parameters are used for condition monitoring and health assessment of the boiler system. The steps for implementing condition monitoring and health assessment of boiler equipment in thermal power units based on the estimation results specifically include: The degree of degradation of the heat transfer coefficient of the water-cooled wall ζ sl Represented as: In the formula, K represents the heat transfer coefficient of the water-cooled wall under rated load when there is no slag buildup inside the furnace. sl The actual heat transfer coefficient of the water-cooled wall under rated load represents the evaluation result; Based on the operating conditions, an evaluation standard for in-furnace slagging characteristics based on the degree of heat transfer coefficient degradation is obtained, expressed as: In the formula, S l and S h The threshold value is pre-calibrated.

Images