A flue gas temperature deviation adjustment system and method for boilers

By applying multi-source data acquisition and collaborative control modules, multi-dimensional collaborative adjustment of boiler flue gas temperature deviation was achieved, solving the problems of single adjustment methods and lag in response in existing systems, and improving the boiler's operational stability and economy.

CN123114151APending Publication Date: 2026-11-17SHANGHAI BOILER WORKS CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611201182.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-10
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing boiler flue gas temperature regulation systems have limited adjustment methods, slow response, and lack of coordinated linkage mechanisms. This results in repeated fluctuations in flue gas temperature deviation, making it difficult to accurately eliminate local hot spots. Furthermore, the adjustment process sacrifices boiler efficiency, affecting safety and economy.

Method used

The system employs a multi-source data acquisition module, a collaborative control module, a flue gas recirculation adjustment module, a secondary air distribution adjustment module, and a burnout air offset adjustment module to form a closed-loop system for data acquisition, decision-making, execution, and interlocking. Through a collaborative control algorithm, the system dynamically allocates adjustment weights and combines flue gas temperature deviation change rate prediction and proactive adjustment strategies to achieve coordinated linkage of multi-dimensional adjustment methods.

Benefits of technology

It effectively eliminates flue gas temperature shocks under transient operating conditions, improves system operation stability and safety, reduces energy consumption, and enhances the boiler's operating economy and environmental friendliness.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123114151A_ABST
    Figure CN123114151A_ABST
Patent Text Reader

Abstract

This invention discloses a flue gas temperature deviation adjustment system and method for boilers, belonging to the field of boiler combustion control technology. It includes a multi-source data acquisition module, a collaborative control module, a flue gas recirculation adjustment module, a secondary air distribution adjustment module, a burnout air bias adjustment module, and an adjustment interlock module. The collaborative control module incorporates a collaborative adjustment algorithm based on multivariate predictive control. This algorithm dynamically allocates the adjustment weights and response priorities of the flue gas recirculation adjustment module, the secondary air distribution adjustment module, and the burnout air bias adjustment module according to the amplitude, rate of change, and spatial distribution characteristics of the flue gas temperature deviation. The adjustment interlock module is used to establish a command conflict detection and arbitration mechanism among the modules. This invention solves the problems of single adjustment methods, slow response, and poor coordination in existing technologies, significantly reducing the amplitude and fluctuation of flue gas temperature deviation, and improving boiler operation safety and economy.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of boiler combustion control technology, specifically to a flue gas temperature deviation adjustment system and method for boilers. Background Technology

[0002] Boiler flue gas temperature deviation is a key technical challenge affecting the operational safety and economy of thermal power units. Existing regulation technologies suffer from the following significant shortcomings: regulation methods are singular and independent, lacking a coordinated linkage mechanism. Regulation systems such as flue gas recirculation, secondary air ratio, and burnout air offset typically operate independently with different control logics. Each system adjusts based solely on its own set thresholds and algorithms, without information sharing or coordination. When flue gas temperature deviation occurs, one system may increase the recirculated flue gas volume to lower the temperature, while another system may decrease the secondary air volume to alter the flame center. These effects can cancel each other out or even excessively overlap, leading to repeated flue gas temperature oscillations. The regulation process cannot converge over a long period, and in severe cases, it can even cause combustion instability.

[0003] Existing systems typically only initiate regulation after the flue gas temperature deviation has already formed and exceeded the alarm threshold. However, boilers exhibit a significant delay between combustion disturbances and flue gas temperature response. When the unit experiences transient conditions such as rapid load increases or decreases, coal mill switching, or soot blowing operations, the flue gas temperature deviation can increase dramatically within a short period. Subsequent regulation cannot suppress the peak deviation in time, often causing localized overheating of heating surfaces and even leading to tube rupture accidents.

[0004] Existing systems primarily address overall deviation adjustments, making it difficult to precisely eliminate localized hot or cold spots within the flue cross-section. Traditional methods, such as adjusting the burnout air swash plate angle or flue gas dampers, can only alter the overall temperature difference between the left and right sides. However, they lack precise control over localized high-temperature zones within the flue cross-section caused by uneven burner output or pulverized coal distribution deviations. Even if the average temperature difference between the left and right sides is improved, localized heated surfaces may still remain in a state of overheating for extended periods.

[0005] The adjustment process often comes at the cost of boiler thermal efficiency. Common methods to reduce localized high temperatures include increasing excess air or increasing flue gas recirculation, but this increases exhaust heat loss and fan energy consumption, and may also lead to higher fly ash carbon content, resulting in a significant decrease in boiler efficiency. Existing systems lack the ability to make multi-objective global optimization decisions among flue gas temperature deviation control, combustion efficiency, and pollutant emissions, making it difficult to achieve a comprehensive optimal balance between safety, economy, and environmental protection.

[0006] Therefore, the present invention aims to provide a flue gas temperature deviation adjustment system and method for boilers to overcome the deficiencies in the prior art. Summary of the Invention

[0007] The purpose of this invention is to provide a flue gas temperature deviation adjustment system and method for boilers, so as to solve the problems of single adjustment means, slow response and poor coordination in the prior art.

[0008] To solve the above-mentioned technical problems, the present invention provides the following technical solution, specifically including:

[0009] A flue gas temperature deviation regulation system for boilers, the system comprising:

[0010] The multi-source data acquisition module, collaborative control module, flue gas recirculation adjustment module, secondary air distribution adjustment module, burnout air offset adjustment module, and adjustment interlock module form a closed-loop operation of acquisition, decision-making, execution, interlocking, and optimization.

[0011] The multi-source data acquisition module is used to collect real-time data on boiler furnace outlet flue gas temperature distribution, flue gas temperature difference across the horizontal flue, SCR inlet flue gas temperature distribution, and boiler operating parameters. These operating parameters include load command, coal type characteristics, total air volume, oxygen content, pulverized coal feed rate to each burner, primary air velocity, secondary air volume, and burnout air volume. Specifically, the furnace outlet flue gas temperature distribution is obtained through a matrix thermocouple array positioned below the screen-type superheater; the flue gas temperature difference across the horizontal flue is obtained through flue gas temperature probes positioned on both sides of the diversion chamber; and the SCR inlet flue gas temperature distribution is obtained through a multi-point armored thermocouple array positioned across the cross-section of the SCR reactor inlet flue.

[0012] The collaborative control module is signal-connected to the multi-source data acquisition module and incorporates a collaborative adjustment algorithm based on multivariate predictive control. This algorithm dynamically allocates the adjustment weights and response priorities of the flue gas recirculation adjustment module, the secondary air distribution adjustment module, and the burnout air bias adjustment module based on the amplitude, rate of change, and spatial distribution characteristics of the smoke temperature deviation. The collaborative control module also includes a smoke temperature deviation rate of change prediction submodule, used to predict the development trend of the smoke temperature deviation within the next 30–300 seconds based on an autoregressive moving average model, and to execute an advance adjustment command when the predicted value exceeds a preset alarm threshold.

[0013] The flue gas recirculation regulation module includes a recirculation flue, a recirculation fan, and a flue gas regulation baffle, which is used to extract low-temperature flue gas from the flue after the economizer or the air preheater and inject it into the upper part of the furnace or the horizontal flue.

[0014] The secondary air distribution and adjustment module includes a secondary air box, secondary air dampers at each level, and an actuator, used to adjust the distribution ratio of secondary air volume for each burner in the boiler.

[0015] The burnout air bias adjustment module includes a burnout air nozzle and its swing actuator or layered baffle, used to adjust the vertical deflection angle and horizontal distribution of the burnout air.

[0016] The adjustment interlock module is used to establish adjustment interlock logic between each adjustment module to avoid command conflicts that could cause smoke temperature oscillations.

[0017] The layered adjustment strategy built into the collaborative control module is as follows: when the flue gas temperature deviation amplitude is less than the first preset threshold, the secondary air distribution adjustment module is adjusted first; when the flue gas temperature deviation amplitude is greater than or equal to the first preset threshold and less than the second preset threshold, the flue gas recirculation adjustment module is activated collaboratively, and the adjustment weight of the secondary air distribution module is dynamically adjusted; when the flue gas temperature deviation amplitude is greater than or equal to the second preset threshold, or the flue gas temperature deviation change rate exceeds the preset rate threshold, the burnout air offset adjustment module is activated simultaneously for deep intervention, and the rapid adjustment action of the secondary air module is temporarily frozen.

[0018] The formula for calculating the adjustment weight of the secondary air distribution and adjustment module is as follows: ;

[0019] The formula for calculating the adjustment weight of the flue gas recirculation adjustment module is as follows: ;

[0020] The formula for calculating the adjustment weight of the burnout air bias adjustment module is as follows: ;

[0021] The smoke temperature deviation change rate prediction submodule adopts the ARMA(2,1) model, and its structure is as follows: ;

[0022] The formula for calculating the advance adjustment amount is: ;

[0023] A method for adjusting flue gas temperature deviation in boilers includes the following steps:

[0024] S1. Real-time acquisition of flue gas temperature distribution data and operating parameters at key locations of the boiler through a multi-source data acquisition module;

[0025] S2. The collaborative control module calculates the amplitude ΔT and the rate of change d(ΔT) / dt of the current smoke temperature deviation, and identifies the dominant area of ​​the deviation based on the spatial distribution characteristics.

[0026] S3. Based on the hierarchical adjustment strategy, dynamically allocate the weights of the three adjustment modules: flue gas recirculation, secondary air distribution, and burnout air bias.

[0027] S4. The prediction submodule is used to predict the trend of smoke temperature deviation in the next 30 seconds based on the ARMA model. If the predicted value exceeds the alarm threshold, the advance adjustment amount is calculated and added to the current deviation.

[0028] S5. Based on the assigned weights and the preset adjustment amount-effect mapping model, calculate the adjustment instructions for each module and send them to the execution mechanism.

[0029] S6. The interlocking module performs conflict detection and arbitration on the instructions of each module to ensure consistent execution logic.

[0030] S7. During the adjustment process, continuously monitor the changing trend of the smoke temperature deviation. If the deviation elimination rate is lower than expected, trigger the incremental learning module to correct the mapping model parameters.

[0031] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0032] This invention achieves coordinated operation of multi-dimensional adjustment methods. Through a hierarchical weight allocation mechanism, it automatically selects the optimal adjustment combination based on the magnitude and rate of change of the deviation, avoiding the problems of over-adjustment or under-adjustment of a single adjustment method. It introduces a predictive and proactive adjustment strategy for the rate of change of flue gas temperature deviation, transforming traditional reactive adjustment into proactive prediction, effectively eliminating flue gas temperature shocks under transient conditions. An interlocking logic for adjustment is established to prevent conflicts between module commands, improving the stability and safety of system operation. The adjustment process also considers operational economy, prioritizing low-energy secondary air adjustment when the deviation is small, significantly reducing the energy consumption of the flue gas recirculation fan and the burnout air system. Attached Figure Description

[0033] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0034] Figure 1 This is a schematic diagram of the module connection of a boiler flue gas temperature deviation adjustment system based on the present invention;

[0035] Figure 2 This is a flowchart of the steps of a boiler flue gas temperature deviation adjustment method based on the present invention. Detailed Implementation

[0036] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 are within the scope of protection of the present invention.

[0037] Please see Figures 1-2 The present invention provides the following technical solution:

[0038] like Figure 1 As shown, this embodiment uses a 600MW subcritical tangential combustion boiler as an example. The following hardware equipment is deployed at the boiler site.

[0039] Eighteen K-type armored thermocouples, arranged in three rows and six columns at equal intervals along the furnace width at the furnace outlet, have a measuring range of 0–1200℃ and an accuracy of ±1.5℃. Two flue gas temperature probes are placed on each of the left and right side walls of the horizontal flue gas turning chamber to calculate the temperature difference between the two sides. Sixteen thermocouples, arranged in a 4×4 grid, are placed in the cross-section of the SCR reactor inlet flue gas duct to detect the SCR inlet flue gas temperature distribution.

[0040] It employs a redundant industrial control computer, configured with an Intel Xeon processor, running a Linux real-time operating system, with a control cycle of 1 second. It includes a built-in collaborative regulation algorithm software package, comprising a data preprocessing module, a deviation feature extraction module, a hierarchical weight allocation module, a regulation calculation module, and an ARMA prediction module.

[0041] Flue gas is drawn from the flue behind the air preheater and sent to the upper part of the furnace below the burnout air layer via a recirculation fan. An electrically adjustable damper is installed on the recirculation flue, with an opening degree of 0–100%.

[0042] There are a total of 24 secondary air dampers for each burner layer, all of which are electrically operated.

[0043] It has 6 burnout air nozzles, each equipped with a horizontal swing actuator and a vertical swing actuator.

[0044] The modules are connected sequentially according to the signal flow direction of "multi-source data acquisition module, collaborative control module, flue gas recirculation adjustment module, secondary air distribution adjustment module, and burnout air bias adjustment module". The adjustment interlock module is connected to the three adjustment modules respectively to arbitrate commands. The prediction submodule obtains historical data from the collaborative control module and returns the advance adjustment amount, forming a complete closed loop from data acquisition to intelligent decision-making to execution interlock.

[0045] Before the system is put into operation, the operator inputs the first preset threshold T1=5℃, the second preset threshold T2=15℃, and the rate of change benchmark value R into the collaborative control module through the human-machine interface. over =0.05℃ / s, secondary wind regulation saturation deviation threshold T sat =10℃, time constant τ=60s, lead gain coefficient K pre =0.8. The proportional coefficient of the adjustment amount-effect mapping model is pre-calibrated by field step test K. sec =0.2% / ℃, K rec =500 m³ / h·℃,K θ =1.0° / ℃.

[0046] like Figure 2 As shown, the implementation process is broken down as follows:

[0047] Step S100: The multi-source data acquisition module acquires the following data at a 1-second interval;

[0048] The flue gas temperature at the furnace outlet at 18:00 is denoted as T. furn [i][j], where i=1~3, j=1~6; the average temperature T of the two cigarettes on the left side of the horizontal flue. left The average temperature T of the two cigarettes on the right right The smoke temperature at the SCR inlet at 16 points is denoted as T. scr [k][l], where k=1~4, l=1~4; operating parameters include load command L (MW), coal feed rate B of each pulverizer. m (t / h), Total air volume Q air (m³ / h), Oxygen content (O2%), Opening degree of each secondary air damper (S) sec [m] (%), recirculated flue gas volume Q rec (m³ / h), burnout wind deflection angle θ H and θ V (°).

[0049] Preprocessing includes: removing obvious erroneous values, such as temperature spikes exceeding 200℃ / s, smoothing with sliding median filtering, and aligning timestamps according to device identifiers.

[0050] Step S200: Calculation of flue gas temperature deviation characteristics. Define the current horizontal flue gas temperature deviation amplitude ΔT as... Where ΔT is the temperature difference (°C) between the two sides of the horizontal flue, T left T is the arithmetic mean (°C) of the measurements taken by the two smoke temperature probes on the left. right This is the arithmetic mean (°C) of the measurements taken by the two smoke temperature probes on the right. To eliminate the influence of instantaneous fluctuations, the exponentially weighted moving average of the five most recent calculations is used as the final deviation. ; where ΔT smooth ΔT(t) is the smoothed deviation value (°C) at time t, ΔT(t) is the instantaneous deviation value (°C) at time t, β is the smoothing coefficient, taken as 0.3, ΔT smooth (t-1) represents the smoothing deviation value (°C) of the previous time step.

[0051] The rate of change of flue gas temperature deviation, dΔT / dt, is calculated by backward difference. Where dΔT / dt is the rate of change of smoke temperature deviation (°C / s), Δt is the sampling interval, taken as 1 second, and ΔT smooth (t-Δt) is the smoothing deviation value (°C) at the previous sampling time.

[0052] Spatial distribution characteristics are identified by comparing the deviation of the flue gas temperature at 18 points at the furnace outlet from the design average value. If the temperature on the left side of the furnace is generally higher than that on the right side, it is determined to be "overheating on the left side"; otherwise, it is "overheating on the right side". If the temperature in the central area is higher than that on both sides, it is "center offset".

[0053] Taking a certain moment as an example, the measured T left =152℃, T right =140℃, then ΔT=12℃. ΔT is calculated from the most recent 5 smoothed values. smooth =12℃. The rate of change in 1 second before and after is dΔT / dt=0.08℃ / s, and the spatial distribution is "overheating on the left".

[0054] Step S300: Layered Adjustment Strategy and Weight Allocation. The collaborative control module performs layered judgment based on the current deviation amplitude and rate of change. Assume that under the current operating condition, ΔT... smooth =12℃, dΔT / dt=0.08℃ / s. Because ΔT smooth =12℃ is greater than or equal to T1=5℃ and less than T2=15℃, but the rate of change is 0.08℃ / s greater than R. over =0.05℃ / s, according to the stratification strategy, enter the "Start burnout wind offset and freeze secondary wind rapid adjustment" branch.

[0055] Calculate the weights of each module: ;W sec Assign weights to the secondary air control module, ΔT = 12℃, T sat =10℃, 1-12 / 10=-0.2 (in parentheses), take 0, therefore W sec =0. α(t)=e −t / τ Here, W is the attenuation coefficient. sec It is already 0, so no specific value is needed.

[0056] Among them, W over For the weight of the burnout air bias adjustment module, |dΔT / dt| = 0.08℃ / s, R over =0.05℃ / s, then W over =min(1,0.08 / 0.05)=min(1,1.6)=1.

[0057] Among them, W rec Substituting the weights of the flue gas recirculation regulation module into the equation, we get W. rec =1-0-1=0. At this point, deep intervention is entirely achieved by the burnout air bias, while secondary air and recirculation are temporarily not involved. If, under another operating condition, ΔT=8℃, dΔT / dt=0.02℃ / s, and the adjustment has been in place for 20 seconds, then α(t)=e −20 / 60 =e −0.3333 ≈0.7165. W sec =max(0,1-8 / 10)·0.7165=0.2×0.7165=0.1433, W over =min(1,0.02 / 0.05)=0.4, W rec=1-0.1433-0.4=0.4567, at this point the three modules are adjusted in a coordinated manner according to the weights 0.1433:0.4567:0.4.

[0058] Step S400: Advanced predictive control. The predictive submodule adopts the ARMA(2,1) model, and its structure is as follows:

[0059] ;

[0060] Where y(t) is the smoke temperature deviation (°C) at time t. θ is the autoregressive coefficient, θ1 is the moving average coefficient, and ε(t) is the white noise sequence following a normal distribution N(0, σ). 2 ).

[0061] Parameter identification is performed online using the smoke temperature deviation sequence {y(1),y(2),…,y(300)} from the most recent 300 sampling points as the training set. The recursive least squares method is used for online identification. The information vector φ(t) = [y(t-1),y(t-2),ε(t-1)] is defined. T parameter vector Recurrence formula:

[0062] ;

[0063] ;

[0064] in, P(t) is the parameter estimation vector at time t, and P(t) is the covariance matrix. Initialize P(0) = 10. 6 I, where I is the identity matrix. (0) = [0,0,0] T The model parameters are updated each time a new data point is received. After 300 recursive points, the current model parameters are obtained, for example... , , .

[0065] By recursively calculating 30 steps backward from the current time t, we obtain the predicted deviation value y for the next 30 seconds. pred (t+30). Single-step prediction recursive formula: ,in, (t+k|t) represents the predicted value for the future time t+k based on the information at time t. For j>0, (t+j|t)=0, initial value (t|t)=y(t), (t-1|t)=y(t-1). Recursively applying this to k=30, we obtain y. pred (t+30).

[0066] Let the current deviation be y(t) = 8℃. The result of recursive calculation is y. pred (t+30)=11℃. Lead gain coefficient K pre =0.8, then the advance adjustment amount is: =0.8×(11-8)=2.4℃. Where, ΔU pre This is the advance adjustment amount (°C). This advance amount is added to the current deviation control input, and the controller considers the current effective deviation to be 8 + 2.4 = 10.4°C, thus increasing the adjustment intensity in advance. The predictive model is updated every 30 seconds to ensure adaptability to time-varying operating conditions.

[0067] Step S500: Adjustment Calculation. Specific instructions are calculated based on the weights of each module and the adjustment-effect mapping model. This step uses ΔT = 8℃, dΔT / dt = 0.02℃ / s, and W... sec =0.1433, W rec =0.4567, W over Taking the operating condition with a value of 0.4 as an example, the target deviation ΔT target =2℃, current advance adjustment amount ΔU pre =2.4℃, then the current effective deviation is ΔT current +ΔU pre =10.4℃. The controller will calculate adjustment commands based on the effective deviation to suppress the trend of future deviation increase in advance.

[0068] Secondary air regulation volume , where ΔS sec K represents the change in the opening degree of the secondary damper (%). sec The secondary wind proportionality coefficient is taken as 0.2% / ℃, ΔT target =2℃, W sec =0.1433. Substituting this into the equation, we get ΔS. sec =0.2×(2-10.4)×0.1433=-0.2407%, the negative sign indicates a reduction in opening. The actual result is rounded down to -0.24%.

[0069] Flue gas recirculation adjustment amount , where ΔQ rec K represents the change in recirculated flue gas volume (m³ / h). rec The rate of change of recirculated air volume required per unit deviation is taken as 500 m³ / h·℃, W rec =0.4567. Substituting this into the equation, we get ΔQ. rec =500×(2-10.4)×0.4567=-1918.1m 3 / h, the negative sign indicates a reduction in the amount of recirculated flue gas. The current base value is 8000 m³ / h, the target value is 6084.9 m³ / h, and the slope change rate is limited to 200 m³ / h / s.

[0070] Burnout air offset adjustment amount , where Δθ V K represents the change in the vertical oscillation angle of the burnout wind (°). θ Let T be the rate of change of the vertical swing angle required per unit deviation, taken as 1.0° / ℃. left -T right The sign function is T, which takes +1 when the smoke temperature on the left is higher than that on the right, and -1 otherwise. left >T right Therefore, we take +1, W over =0.4. Substituting this into the equation, we get Δθ. V =1.0×(2-10.4)×(+1)×0.4=-3.36° The negative sign indicates a downward swing. Due to the limitation of the range of motion, -3.0° is taken in actual execution.

[0071] Step S600: The interlock module is adjusted to detect and arbitrate the above commands. A directional consistency check is performed; the current deviation is left-side overheating, and all commands should reduce the left-side smoke temperature. The secondary air command reduces the left-side secondary air volume, the recirculation command reduces the total recirculation volume, and the burnout air swings downwards to lower the flame center, resulting in a more uniform smoke distribution and correct direction. No conflict.

[0072] Checking the priority, assuming the burnout air command has just been issued, we enter a 15-second action dead zone. At this time, the rapid adjustment command for secondary air is frozen, i.e., ΔS sec Do not execute for now, maintain the opening degree from the previous moment; multiply the loop instruction by 0.5, which is the actual execution. =0.5×(-1918.1)=-959.05m 3 / h. The burnout wind command was executed normally.

[0073] Finally, check the amplitude limit. =959.05m 3 / h is less than 2000 m³ / h, which is acceptable; Δθ V =3.36° exceeds the 3° limit, so the actual execution is -3.0°, which is acceptable; the secondary wind does not operate due to freezing, so there is no amplitude issue. The interlock module sends the arbitrated instructions to each executing agency.

[0074] Step S700: Incremental learning correction. After adjustment, record the deviation before and after adjustment: ΔT before =8℃, ΔT after =3℃, adjustment time Δt 调节 =120 seconds. Actual deviation elimination rate: , where r actual Let r be the actual deviation elimination rate (°C / s). Substituting this into the equation, we get r. actual =(8-3) / 120=0.0417℃ / s.

[0075] Theoretical expected rate, ℃ / s, relative deviation is |r actual -r exp ∣ / r exp =∣0.0417-0.05∣ / 0.05=0.166=16.6%, which is less than the 20% threshold, so no correction is triggered. If the deviation exceeds 20%, then the recursive least squares correction of the mapping model coefficients is initiated.

[0076] If the actual elimination rate is significantly lower than expected, it indicates that the current K rec Too small. New observation data: ΔT err =6℃, actual adjustment ΔQ rec =-1918.1m 3 / h. Update K rec : , , where K rec_new K is the updated scaling factor. rec_old =500, P i-1 The previous covariance value is given, λ = 0.98 is the forgetting factor, and ΔT is the value of the previous covariance. err =6℃, ΔQ rec =-1915.1m 3 / h. Assume P i-1 =0.01, then the denominator 0.98 + 0.01 × 36 = 0.98 + 0.36 = 1.34. P i =0.01 / 1.34≈0.00746. Substitute K rec_new =500 + 0.00746 × 6 × (1918.1 - 500 × 6) = 500 - 48.45 = 451.55. The new coefficient is approximately 451.0 m³ / h·℃, which will be used for subsequent adjustments. This learning process is fully automated, and the parameters are updated every 20 sets of valid data.

[0077] After implementing this system on a 600MW unit, a three-month comparative test was conducted. During the test, the boiler load fluctuated frequently between 300MW and 600MW, and the coal mill was switched an average of three times per day. Compared with the control group, the average maximum flue gas temperature difference in the test group decreased from 22.5℃ to 6.8℃, an improvement of 69.8%; the percentage of times the flue gas temperature difference exceeded 15℃ decreased from 38.2% to 5.1%; the number of times the reheater wall temperature exceeded the limit decreased from 12 times per month to 2 times; the average daily power consumption of the recirculation fan decreased from 1240kWh to 890kWh; and the average NOx emission decreased from 185mg / m³ to 172mg / m³. The unit's operational safety, economy, and environmental performance were all significantly improved.

[0078] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0079] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A flue gas temperature deviation adjusting system applied to a boiler, characterized by, include: Multi-source data acquisition module, collaborative control module, flue gas recirculation adjustment module, secondary air distribution adjustment module, burnout air offset adjustment module, and adjustment interlock module; The multi-source data acquisition module is used to collect real-time data on boiler furnace outlet flue gas temperature distribution, flue gas temperature difference on both sides of the horizontal flue, SCR inlet flue gas temperature distribution, and boiler operating parameters. The collaborative control module incorporates a collaborative adjustment algorithm based on multivariable predictive control. The collaborative adjustment algorithm dynamically allocates the adjustment weights and response priorities of the flue gas recirculation adjustment module, the secondary air distribution adjustment module, and the burnout air bias adjustment module according to the amplitude, rate of change, and spatial distribution characteristics of the flue gas temperature deviation. The collaborative control module is also equipped with a smoke temperature deviation change rate prediction submodule, which is used to predict the development trend of smoke temperature deviation based on an autoregressive moving average model, and to execute an advance adjustment command when the predicted value exceeds a preset alarm threshold. The flue gas recirculation regulation module is used to extract low-temperature flue gas from the flue after the economizer or the air preheater and inject it into the upper part of the furnace or the horizontal flue. The secondary air distribution adjustment module is used to adjust the distribution ratio of secondary air volume of each burner in the boiler; the burnout air offset adjustment module is used to adjust the vertical deflection angle and horizontal distribution of burnout air; the adjustment interlock module is used to establish adjustment interlock logic between each adjustment module to avoid command conflicts that cause flue gas temperature oscillation.

2. A flue gas temperature deviation adjusting system for a boiler according to claim 1, wherein The layered adjustment strategy built into the collaborative control module is as follows: when the flue gas temperature deviation amplitude is less than the first preset threshold, the secondary air distribution adjustment module is adjusted first; when the flue gas temperature deviation amplitude is greater than or equal to the first preset threshold and less than the second preset threshold, the flue gas recirculation adjustment module is activated collaboratively, and the adjustment weight of the secondary air distribution module is dynamically adjusted; when the flue gas temperature deviation amplitude is greater than or equal to the second preset threshold or the flue gas temperature deviation change rate exceeds the preset rate threshold, the burnout air bias adjustment module is activated simultaneously, and the rapid adjustment action of the secondary air distribution module is temporarily frozen.

3. The flue gas temperature deviation adjustment system for boilers according to claim 1, characterized in that, The formula for calculating the adjustment weight of the secondary air distribution and adjustment module is as follows: Where ΔT is the current smoke temperature deviation value, T sat The saturation deviation threshold for secondary air regulation is α(t), which is a weighting coefficient that decays over time. The formula for calculating the regulation weight of the flue gas recirculation regulation module is as follows: The formula for calculating the adjustment weight of the burnout air bias adjustment module is as follows: Where d(ΔT) / dt is the rate of change of smoke temperature deviation, R over This is the benchmark value for the rate of change.

4. The flue gas temperature deviation adjustment system for boilers according to claim 1, characterized in that, The model structure used in the smoke temperature deviation change rate prediction submodule is as follows: Where y(t) is the time series of smoke temperature deviation, θ1 is the autoregressive coefficient, θ1 is the moving average coefficient, and ε(t) is white noise; Model parameters were identified online using the least squares method based on the most recent 300 sampling points; The formula for calculating the advance adjustment amount is: K pre y is the lead gain coefficient. pred (t+Δt) represents the prediction deviation.

5. A flue gas temperature deviation adjustment system for boilers according to claim 1, characterized in that, The adjustment interlock module includes action priority rules: during the execution of the burnout air bias action, the rapid adjustment command of the secondary air distribution module is frozen, and the adjustment amount of the flue gas recirculation module is halved; directional consistency rules: if the current flue gas temperature deviation is overheating on the left, the adjustment direction of all three modules must reduce the flue gas temperature on the left, otherwise the weight is reset to zero and an alarm is triggered; and action amplitude limits: within a single control cycle, the changes in recirculated flue gas volume, the opening degree of the secondary air damper, and the oscillation angle of the burnout air shall not exceed their respective preset thresholds.

6. The flue gas temperature deviation adjustment system for boilers according to claim 1, characterized in that, In the multi-source data acquisition module, the flue gas temperature distribution at the furnace outlet is obtained through a matrix thermocouple array arranged below the screen-type superheater; the flue gas temperature difference on both sides of the horizontal flue is obtained through flue gas temperature probes arranged on both sides of the turning chamber; and the flue gas temperature distribution at the SCR inlet is obtained through a multi-point armored thermocouple array arranged on the cross-section of the SCR reactor inlet flue. The operating parameters include load command, coal type characteristics, total air volume, oxygen content, pulverized coal feed rate of each burner, primary air velocity, secondary air volume, and burnout air volume.

7. A flue gas temperature deviation adjustment system for boilers according to claim 1, characterized in that, The system also includes a regulation-effect mapping model, which is established through field step tests. The mapping relationship of the secondary air distribution regulation module is as follows: K sec This is the proportionality coefficient; The mapping relationship of the flue gas recirculation regulation module is as follows: K rec The rate of change of recirculated air volume required per unit deviation; the mapping relationship of the burnout air offset adjustment module is as follows: K θ The rate of change of the vertical swing angle required per unit deviation.

8. A flue gas temperature deviation adjustment system for boilers according to claim 1, characterized in that, The system is also equipped with an incremental learning module, which compares the actual deviation elimination rate with the theoretical expected rate after adjustment. If the relative deviation exceeds a preset threshold, a recursive least squares correction is initiated to update the proportional coefficient in the adjustment amount-effect mapping model. The incremental learning module performs a parameter update once for every preset number of valid data accumulated, and uses a preset forgetting factor.

9. A flue gas temperature deviation adjustment system for boilers according to claim 1, characterized in that, The collaborative control module adopts a redundant industrial control computer; the flue gas recirculation adjustment module includes a recirculation flue, a variable frequency recirculation fan, and an electric adjustment damper; the secondary air distribution adjustment module includes a secondary air box and electric actuators for each layer of secondary air dampers; the burnout air offset adjustment module includes burnout air nozzles and their horizontal and vertical swing actuators.

10. A method for adjusting flue gas temperature deviation in boilers, characterized in that, Includes the following steps: S1. Real-time acquisition of flue gas temperature distribution data and operating parameters at key locations of the boiler through a multi-source data acquisition module; S2. The collaborative control module calculates the amplitude ΔT and the rate of change d(ΔT) / dt of the current smoke temperature deviation, and identifies the dominant area of ​​the deviation based on the spatial distribution characteristics. S3. Based on the hierarchical adjustment strategy, dynamically allocate the weights of the three adjustment modules: flue gas recirculation, secondary air distribution, and burnout air bias. S4. The prediction submodule is used to predict the development trend of smoke temperature deviation based on the model. If the predicted value exceeds the alarm threshold, the advance adjustment amount is calculated and added to the current deviation. S5. Based on the assigned weights and the preset adjustment amount-effect mapping model, calculate the adjustment instructions for each module and send them to the execution mechanism. S6. The interlocking module performs conflict detection and arbitration on the instructions of each module to ensure consistent execution logic. S7. During the adjustment process, continuously monitor the changing trend of the smoke temperature deviation. If the deviation elimination rate is lower than expected, trigger the incremental learning module to correct the mapping model parameters.