Boiler optimization control method and system adaptive to coal type change

Abstract

The invention discloses a boiler optimization control method and system self-adaptive to coal type changes, and belongs to the technical field of boiler intelligent control, and the method comprises the steps: building a coal type characteristic factor set through collecting coal quality on-line monitoring data; judging a current coal type through a pre-trained coal type recognition model; extracting matched control parameters from a boiler control parameter database based on the coal type, and performing adaptive correction in combination with the boiler operation state parameters to obtain a target control parameter set; the parameter set is loaded to a boiler control system, and executing mechanisms such as a combustor and an air supply system are adjusted in a linkage mode; meanwhile, continuously monitoring coal quality change trend and operation feedback data, and automatically triggering control strategy reconstruction if coal type sudden change or thermal deviation abnormity is detected; the method has good real-time performance, self-adaptability and stability, and the operation safety of the boiler under the scene of multi-coal mixed combustion or frequent coal change can be remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent boiler control, in particular to a boiler optimization control method and system adaptive to coal type change. BACKGROUND

[0002] With the diversification of energy structure, industrial boilers face the realistic demand of mixed burning of multiple coal types or frequent replacement of coal types during operation, especially in the fields of cogeneration and industrial steam supply. Due to the significant differences in volatile matter, ash melting point, sulfur content, calorific value and combustion characteristics of different coal types, the conventional boiler control system using fixed control parameters is difficult to adapt to such dynamic changes, which easily leads to problems such as decreased boiler combustion efficiency, increased pollutant emissions, serious slagging, load fluctuation, etc.

[0003] For example, when the boiler is switched from lignite to bituminous coal, due to the lower volatile matter and higher ignition point of bituminous coal, if the air supply ratio, combustion air distribution area and combustion chamber temperature control strategy are not adjusted in time, it is easy to cause the flame center to move backward, the boiler thermal deviation to increase, and even the problems of furnace slagging and even flameout trip. At the same time, frequent human intervention not only increases the operation and maintenance cost, but also is difficult to realize real-time response to coal type change. SUMMARY

[0004] The purpose of the present application is to provide a boiler optimization control method and system adaptive to coal type change to solve the problems in the background art.

[0005] In order to achieve the above purpose, the present application provides the following technical scheme: a boiler optimization control method adaptive to coal type change, comprising: S100, collecting current fuel feeding parameters of the boiler, including online monitoring data of coal quality, obtaining a set of coal type characteristic factors M, including a characteristic vector composed of volatile matter, ash content, sulfur content and low calorific value indexes of the coal type; S200, inputting the set of coal type characteristic factors M into a pre-trained coal type recognition model to recognize the coal type L corresponding to the current combustion coal type; S300, according to the coal type L, retrieving the optimal control parameter set P matched with the coal type L from a boiler control parameter database; S400, combining the current boiler operation state parameters R to correct the control parameter set P to obtain a target control parameter set P' adaptive to the current operating condition; wherein R includes boiler load, main steam pressure, flue gas temperature and furnace negative pressure data; S500, applying the target control parameter set P' to the boiler control and dynamically adjusting the combustion process; S600, continuously monitor the coal quality change trend and the boiler operation feedback data, if the coal type changes or the boiler thermal deviation increases, re-identify and adjust the control strategy.

[0006] Preferably, the S200 comprises: S201, the obtained coal type characteristic factor set M is normalized by the minimum-maximum standardization method to normalize the characteristic data of different dimensions; S202, input the normalized characteristic data into the pre-trained coal type identification model, the model adopts a hybrid structure of convolutional neural network and support vector machine, wherein the convolutional layer is used to extract the deep nonlinear correlation of coal type characteristics, and the support vector machine is used to realize multi-classification decision; S203, according to the classification confidence vector output by the model, the confidence of the identification result is weighted and judged, when the highest confidence is lower than the preset threshold, the approximate characteristic sample is called for secondary matching; S204, output the coal type L with the highest identification confidence.

[0007] Preferably, the S300 comprises: S301, receive the output coal type L and input it into the boiler control parameter database as a query keyword; S302, match the preset control parameters in the database according to the coal type L, including the primary air and secondary air ratio, the burner inclination angle, the air distribution door opening degree and the main combustion zone temperature set value; S303, according to the current boiler load level , the adjustable items in the matched parameter template are adaptively filtered and weighted adjusted to generate a control parameter candidate set which is double adapted to the coal type L and the load state ; S304, replay and compare the operation effects of the parameter combinations in under the same coal type and load conditions, and select the combination with the optimal thermal efficiency and stable emission as the final control parameter set P.

[0008] Preferably, the S400 comprises: S401, collect the current operation state parameters R of the boiler, including the boiler load, the main steam pressure, the exhaust gas temperature and the furnace negative pressure data; S402, construct an operation deviation vector ΔR based on the parameters R, calculate the difference between the current state and the target control boundary to obtain the dynamic deviation of each control index; S403, input the control parameter set P and the operation deviation vector ΔR into the fuzzy adaptive adjustment model; S404, output the corrected target control parameter set P'.

[0009] Preferably, the S500 comprises: S501, input the output target control parameter set P' into the boiler distributed control system DCS; S502, according to the control instructions in P', linkage adjustment of the working state of the burner, air supply system and air distribution door and other actuators is carried out to realize real-time adjustment of the air volume ratio, flame center position and furnace temperature.

[0010] Preferably, the S600 comprises: S601, obtain the coal quality characteristic parameters, and establish a coal quality change trend model by combining a time series analysis method; S602, synchronously collect the boiler operation feedback data, including oxygen content fluctuation, furnace temperature difference and flue gas temperature deviation, and calculate the thermal deviation index T; S603, set the coal type change determination threshold L and the thermal deviation threshold Tt, when the coal quality trend model identifies that the coal type has a sudden change or the thermal deviation index T exceeds Tt, the control strategy reconstruction process is triggered; S604, return to step S100 to re-collect the coal type characteristic factor set M.

[0011] The application also provides a boiler optimization control system for self-adapting to coal type change, comprising: a coal quality perception module, which collects the current fuel feeding parameters of the boiler, including online monitoring data of the coal quality, and obtains a coal type characteristic factor set M, including a characteristic vector composed of the volatile matter, ash content, sulfur content and low calorific value of the coal type; a coal type identification module, which inputs the coal type characteristic factor set M into a pre-trained coal type identification model to identify the coal type L corresponding to the current combustion coal type; a control parameter matching module, which retrieves the optimal control parameter set P matched with the coal type L from a boiler control parameter database according to the coal type L; a working condition self-adapting correction module, which corrects the control parameter set P in combination with the current boiler operation state parameter R to obtain a target control parameter set P' adapted to the current operation working condition; wherein R includes the boiler load, main steam pressure, flue gas temperature and furnace negative pressure data; a boiler control execution module, which applies the target control parameter set P' to the boiler control and dynamically adjusts the combustion process; a reconstruction triggering module, which continuously monitors the coal quality change trend and the boiler operation feedback data, and re-identifies and adjusts the control strategy if the coal type changes or the boiler thermal deviation increases.

[0012] In the above technical solution, the application provides the technical effects and advantages: 1、The present application realizes the full-process self-adaptive optimization control of the boiler to the change of multiple coal types by constructing a closed-loop control system composed of coal quality perception, coal type identification, parameter matching, working condition correction, control execution and dynamic monitoring. Compared with the traditional mode relying on artificial experience or fixed control logic, the present application can perceive the coal type attributes and the boiler running state in real time, dynamically adjust the control strategy based on intelligent algorithm, significantly improve the combustion efficiency and system response speed, and reduce the risks of slagging, flameout and emission exceeding the standard caused by sudden change of coal type.

[0013] 2、The thermal deviation index modeling and coal quality trend identification mechanism proposed in the present application enables the boiler control system to actively identify coal quality disturbance and running abnormalities, can automatically trigger the identification and regulation process without relying on artificial intervention, and guarantees the stability and intelligent level of long-term operation of the boiler. The overall scheme has high practicality, expandability and engineering deployment value, and is suitable for the intelligent upgrading of coal-fired boilers and the fine control requirements in the multi-coal type blending combustion scene. BRIEF DESCRIPTION OF DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments described in the present application, and other drawings can also be obtained by those skilled in the art based on these drawings.

[0015] Figure 1 The method flowchart of the present application.

[0016] Figure 2 The system module flowchart of the present application. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical scheme and advantages of the embodiments of the present application more clear, the technical scheme in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0018] Embodiment 1, please refer to Figure 1 The self-adaptive coal type change boiler optimization control method described in the present embodiment comprises: S100, collect the current fuel feeding parameters of the boiler, including the online monitoring data of coal quality, obtain a set of coal type characteristic factors M, including a characteristic vector composed of the volatile matter, ash content, sulfur content and low calorific value indexes of the coal type; S200, input the coal type characteristic factor set M into the pre-trained coal type identification model, and identify the coal type L corresponding to the current combustion coal type; S300, according to the coal type L, retrieve the optimal control parameter set P matched with the coal type L from the boiler control parameter database; S400, combine the current boiler operation state parameter R to correct the control parameter set P, and obtain the target control parameter set P' suitable for the current operation condition; wherein R includes boiler load, main steam pressure, flue gas temperature and furnace negative pressure data; S500, apply the target control parameter set P' to the boiler control, and dynamically adjust the combustion process; S600, continuously monitor the coal quality change trend and boiler operation feedback data, if the coal type changes or the boiler thermal deviation increases, re-identify and adjust the control strategy.

[0019] In the present application, first, the real-time parameter acquisition of the fuel coal used by the boiler is needed to obtain the core characteristic index of the coal type. Specifically, step S100 includes online monitoring of the coal quality parameters under the current fuel feeding state of the boiler, and then constructing the coal type characteristic factor set M.

[0020] The coal type characteristic factor set M is used to represent the combustibility, ash composition and energy release characteristics of the current coal type, and its composition includes but is not limited to the following items: Volatile matter (Volatile Matter): represents the ability of coal to decompose and produce combustible gas at high temperature, which is an important parameter to determine the flammability and flame propagation speed of coal; Ash content (Ash Content): reflects the non-combustible part left after coal combustion, which is closely related to the slagging tendency and slagging load; Sulfur content (Sulfur Content): is a key indicator affecting the load of the boiler desulfurization system and the emission control strategy, especially in the scene with high environmental protection requirements; Lower heating value (Lower Heating Value, LHV): is a direct indicator to evaluate the energy released by coal combustion, which has a direct impact on boiler load regulation and fuel economy.

[0021] The above parameters can be collected in real time by the coal quality online analyzer arranged in the coal conveying system or the front section of the coal mill. The online analyzer uses neutron activation, X-ray fluorescence or near-infrared spectroscopy technology for non-destructive testing of coal samples, and the collection frequency can reach once per minute, ensuring that the change of coal quality during the operation of the boiler can be sensed in time.

[0022] After the coal quality data collection is completed, the obtained parameters are uniformly coded to form a feature vector M=[V, A, S, H], where V is the volatile matter, A is the ash content, S is the sulfur content, and H is the low heat value.

[0023] In the present application, the main goal of step S200 is to accurately identify the type L corresponding to the current combustion coal type according to the coal type feature factor set M obtained in step S100. To ensure the recognition accuracy and model generalization ability, step S200 includes the following specific sub-steps: To eliminate the influence of the numerical dimension difference between different coal quality parameters on the model recognition effect, each index in the coal type feature factor set M is uniformly normalized. Specifically, the minimum-maximum normalization method is used, that is, for each feature value X, the minimum value X_min and the maximum value X_max corresponding to it in the historical coal quality data are used for linear transformation, and the normalization expression is: ; wherein X' is the normalized feature value. This method can normalize all input features to the numerical interval [0, 1], thereby enhancing the model's comprehensive processing ability of each dimension feature.

[0024] The normalized coal type feature factor vector M' is input into the pre-trained coal type recognition model. The model uses a hybrid structure combining convolutional neural network (CNN) and support vector machine (SVM), which has strong non-linear feature extraction and high-dimensional classification ability. The specific implementation steps are as follows: The input layer of the CNN network is constructed to receive the M' vector (e.g., containing volatile matter, ash content, sulfur content, and low heat value, a total of 4 dimensions); the convolution layer design includes 2 layers, each layer is set with 16 convolution kernels, the kernel size is 1×2, and the activation function is the rectified linear unit (ReLU), which is used to extract the high-order nonlinear combination relationship of coal type features; on the basis of the feature map output by the convolution network, a fully connected layer is added and a feature embedding vector Z is output; the feature vector Z is input into the support vector machine (SVM) classifier, and the radial basis function (RBF) kernel is used to realize multi-classification decision, and a classification probability (confidence) vector of the current sample under each known coal type category is output.

[0025] The recognition model outputs a classification confidence vector , where each represents the probability of the current input feature corresponding to the i-th coal type category. The category with the highest confidence is selected as the preliminary recognition result.

[0026] If the maximum confidence value is lower than the preset threshold (for example, If the similarity between the current feature sample and the training sample is less than 0.75, it is considered that the similarity between the current feature sample and the training sample is insufficient, and there is a risk of misjudgment. At this time, the system automatically calls the approximate sample library to retrieve the top K historical feature samples (e.g., K=3) with the smallest Euclidean distance to the current feature vector M', and performs a weighted voting secondary discrimination based on their actual coal type labels to improve the recognition stability. The specific voting strategy is as follows: each approximate sample is weighted according to the reciprocal of the distance (i.e., the closer the distance, the higher the weight), and the final coal type identification result is obtained after summarizing.

[0027] After completing the initial identification or secondary matching correction, the final identified coal type L is output, such as "bituminous coal", "anthracite", "lignite" or "lean coal".

[0028] In this invention, the main objective of step S300 is to retrieve a set of control parameters suitable for the current coal type and taking into account the boiler load status, based on the identified coal type L. This step includes the following sub-steps: In this step, the coal type L output from step S204 is first received as the core retrieval keyword for the control strategy invocation. The coal type L is a standard coal classification label, such as "bituminous coal," "anthracite," "lignite," or "lean coal." This keyword is input into the boiler control parameter database constructed in this invention, which pre-stores typical operating parameter templates corresponding to multiple coal types.

[0029] The boiler control parameter database adopts a structured relational storage design, and is optimized with multi-field indexes based on dimensions such as coal type, boiler type, and operating load level, supporting fast matching and combined queries.

[0030] The database is used to query the corresponding control parameter template based on coal type L. Each coal type template contains multiple preset control parameter items, including but not limited to: the primary air to secondary air supply ratio (e.g., 40:60 or 30:70); burner tilt angle (unit: degrees, e.g., ...). ); Air distribution damper opening (percentage setting, controls airflow); Main combustion zone temperature setpoint (unit: degrees Celsius, typically 100 degrees Celsius). The above parameter templates are constructed from historical operating experience, simulation data, and combustion test results, and are customized for the combustion characteristics of different coal types to form a standardized set of initial control parameters.

[0031] To improve the adaptability of the control strategy to real-time operating conditions, the current operating load level R1 of the boiler is introduced (for example, divided into three levels: low, medium, and high, corresponding to load ranges of 0-30%, 30-70%, and 70-100%, respectively). The load level... Together with the coal type L, these parameters are used as joint inputs, and the adjustable parameters are adaptively adjusted within the matched parameter template. Specific adjustment methods include: Introducing load weighting factor ,according to The assigned level is determined; the parameter values ​​in the template are compared with... Joint modeling to form parameter adjustment functions ,in These are the original template parameters. The adjustment coefficient is based on the load; the final output contains a candidate set of multiple combined parameters. Candidate set Includes several groups related to coal type L and current load. The dual-adaptive control parameter combination provides greater operational specificity.

[0032] right The historical operating effects of each parameter combination under the same coal type and load conditions were simulated and compared. The historical data included core indicators such as boiler thermal efficiency, flue gas temperature, NOx emission concentration, and furnace stability.

[0033] The comparison process employs a weighted comprehensive scoring algorithm, assigning a performance score S to each candidate parameter set. The scoring function is as follows: Where η represents thermal efficiency, E represents the normalized value of pollutant emission concentration, and S t The rating represents the operational stability score, where α, β, and γ are empirically set weighting coefficients (e.g., 0.5, 0.3, and 0.2) used to reflect the importance of different indicators. Based on the rating results, the parameter combination with the best overall performance is selected as the final control parameter set P.

[0034] In this invention, step S400 aims to dynamically correct the control parameter set P retrieved from the database by combining it with the current actual operating state of the boiler, in order to obtain a target control parameter set P′ that better matches the characteristics of the current operating conditions, thereby improving the response accuracy and stability of the control system. Specifically, it includes the following sub-steps: In this step, the system collects key operating parameters of the boiler in real time through operating status sensors deployed at critical parts of the boiler, forming an operating status parameter set R. This parameter set includes, but is not limited to: boiler load (in MW or %), representing the current output power level of the boiler; main steam pressure (in MPa, used to characterize the main steam system load); flue gas temperature (in °C, used to reflect thermal energy utilization and tail heating surface conditions); and furnace negative pressure (in Pa, reflecting furnace sealing and induced draft system operation).

[0035] The acquired real-time operating status parameter R is compared with the system's preset target control boundary. By comparing the results, the deviations of each indicator are calculated, forming the operational deviation vector ΔR. The deviation calculation formula is as follows: Where R(i) represents the i-th real-time running parameter, This represents the target control value for this parameter. Target boundary. Based on historical data statistical modeling or empirical rules, it reflects the range of boiler operating indicators under ideal combustion conditions. The operating deviation ΔR can quantify the gap between the current boiler state and the target operating condition, providing a quantitative basis for parameter correction.

[0036] The initial control parameter set P and the operating deviation vector ΔR are input together into the fuzzy adaptive adjustment model to achieve online optimization and correction of the control parameters. Specifically: Based on expert experience, several IF-THEN fuzzy control rules are preset, such as "if the main steam pressure deviation is positive and the flue gas temperature is too high, then the secondary air ratio should be increased appropriately and the main combustion zone temperature setpoint should be lowered"; Membership functions are used to transform continuous deviation values ​​into fuzzy linguistic variables (such as "small", "medium", "large"), and adjustment instructions are generated through rule-based reasoning. Based on feedback on recent control effects, the weighting factors of fuzzy rules are dynamically optimized to improve the model's adaptability under different operating conditions.

[0037] Through the above mechanism, the model performs fine-tuning of key items in the control parameter set P (such as air distribution ratio, temperature setpoint, etc.) and outputs dynamically corrected control parameters.

[0038] After processing by the fuzzy adaptive model, a new set of control parameters P′ is output. This set of parameters fully considers the dual characteristics of the current coal type L and the real-time operating state R, and has higher combustion adaptability and thermal efficiency.

[0039] In this invention, the main objective of step S500 is to apply the modified target control parameter set P′ to the actual boiler control system, and accordingly dynamically adjust the key actuators in the combustion process to achieve closed-loop response control to changes in coal type and load fluctuations. This step includes the following sub-steps: In this step, the target control parameter set P′ is input into the boiler distributed control system (DCS) via a standard industrial communication interface (such as MODBUS, PROFIBUS, or OPC protocol). The target control parameter set P′ includes multiple control command items after operational status correction, mainly including: the primary air to secondary air supply ratio command (e.g., primary air 38%, secondary air 62%); the burner tilt angle setpoint (e.g., ... To achieve adjustment of the flame center forward and backward; air distribution damper opening (e.g., upper damper 60%, lower damper 40%); target temperature value of the main combustion zone (e.g., 1180°C). ); oxygen content setpoint (e.g., 3.5%).

[0040] Before being input into the DCS, the control parameter set must undergo data format encapsulation processing, be converted into a standard data structure recognizable by the DCS, and then be visualized and safely confirmed through a Human-Machine Interface (HMI). The distribution of this data is equipped with an authorization authentication mechanism to prevent erroneous operations or abnormal control commands from affecting the safe operation of the boiler.

[0041] After receiving the control parameter set P′, the boiler DCS performs coordinated adjustments on the key actuators of the boiler according to the set values ​​of each parameter. Specifically, this includes the following: Burner control: The angle of the burner electric actuator is automatically adjusted according to the tilt angle setting, thereby adjusting the back-and-forth movement of the flame center, optimizing the flame coverage area, and adapting to the ignition characteristics of different coal types; Air supply system control: The speed of the frequency converters of the primary air and secondary air fans are controlled separately to achieve dynamic adjustment of air volume; the air supply volume is controlled by closed-loop feedback from the air volume sensor to ensure consistency with the set ratio; Air distribution damper control: Automatically adjusts the opening of the air distribution dampers in the upper, middle and lower parts of the furnace, and drives the butterfly valve or slide gate valve structure through the actuator for fine adjustment to control air distribution and improve local combustion efficiency; Furnace temperature control: By combining the temperature setpoint with real-time monitoring values, and through the coordinated action of airflow regulation and burner control, the temperature of the main combustion zone is maintained within the target range. The above adjustment process is a closed-loop control system. Relying on real-time feedback of furnace temperature, oxygen content, and flue gas temperature, the DCS compares the setpoint and actual values ​​in real time and adjusts and corrects them according to the proportional-integral-derivative (PID) control algorithm. If the deviation exceeds the limit (for example, the oxygen content deviation exceeds ±0.5%), the system can trigger an alarm or fine-tuning mechanism to ensure stable, safe, and energy-efficient operation of the combustion system.

[0042] In this invention, step S600 aims to continuously monitor coal quality fluctuations and boiler operating status, promptly identify sudden changes in coal type or abnormal combustion states based on an intelligent judgment mechanism, and proactively trigger the reconfiguration of the control process to achieve closed-loop adaptive operation. This step includes the following sub-steps: In this step, key characteristic parameters of incoming coal, including volatile matter, ash content, sulfur content and lower heating value, are continuously collected by online coal quality monitoring devices installed in front of the coal conveyor belt or the coal hopper of the furnace, forming a coal quality dataset M(t) arranged in time series.

[0043] To identify coal type fluctuation trends, the system employs time series analysis, preferably incorporating a moving average (MA) model and exponential smoothing algorithm to construct a coal quality change trend model Q(t), which is then dynamically fitted to the feature vector M(t). The model is defined as follows: Where Q(t) is the coal quality trend vector at the current moment, and s is the smoothing coefficient (e.g., set to 0.2~0.4), used to balance the weights of the current sampling and historical trends. The model can be updated in real time to predict whether the coal quality tends to stabilize or undergoes drastic changes.

[0044] Simultaneously collect multiple feedback parameters during boiler operation, including: oxygen content fluctuation range. : Represents the maximum and minimum difference in oxygen content sensor output per unit time; furnace temperature difference : Indicates the maximum temperature difference between different monitoring points in the furnace; flue gas temperature deviation. This indicates the degree of deviation between the actual flue gas temperature and the set target value. Based on the above feedback parameters, a thermal deviation index T is constructed to quantitatively reflect the boiler's operational stability. The calculation formula is: ;in, These are weighting coefficients, set based on operational experience, for example, 0.3, 0.4, and 0.3, to reflect the contribution of each indicator to the system's thermal deviation. This index has real-time updating capabilities and can accurately reflect whether the combustion system exhibits unsteady-state characteristics.

[0045] To automatically detect sudden changes in coal type or abnormal thermal deviations, the system presets two thresholds: Coal type change judgment threshold L: used to determine the coal quality trend model Q(t) and the mean of the previous stable period. Does the Euclidean distance exceed the threshold L? If it does... If so, it is considered that a sudden change has occurred in the coal type; Thermal deviation threshold Used to determine whether boiler operation exceeds the stable range; if the thermal deviation index... This indicates that there is a significant deviation in the combustion system.

[0046] When any of the conditions are met, the control strategy reconfiguration process will be triggered, including operations such as coal type re-identification, control parameter update, and model weight adjustment.

[0047] Once it is determined that the control strategy needs to be reconstructed, immediately return to step S100, re-collect the latest coal quality data, construct a new coal type characteristic factor set M, and re-execute steps S200 to S500 to achieve closed-loop re-identification and strategy update under coal type change or thermal deviation anomaly.

[0048] Example 2, please refer to Figure 2 As shown in this embodiment, an adaptive coal type change-adaptive boiler optimization control system includes: The coal quality sensing module collects the current fuel feed parameters of the boiler, including online coal quality monitoring data, and obtains the coal type characteristic factor set M, which includes a feature vector composed of volatile matter, ash content, sulfur content, and lower heating value of the coal type. The coal type identification module inputs the coal type feature factor set M into the pre-trained coal type identification model to identify the coal type L corresponding to the current burning coal type. The control parameter matching module retrieves the optimal control parameter set P that matches the coal type L from the boiler control parameter database. The adaptive correction module, in conjunction with the current boiler operating status parameter R, corrects the control parameter set P to obtain the target control parameter set P' adapted to the current operating conditions; where R includes boiler load, main steam pressure, flue gas temperature and furnace negative pressure data; The boiler control execution module applies the target control parameter set P' to the boiler control and dynamically adjusts the combustion process; The system reconstructs the trigger module, continuously monitors coal quality changes and boiler operation feedback data, and re-identifies and adjusts the control strategy if the coal type changes or the boiler thermal deviation increases.

[0049] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A boiler optimization control method that adapts to changes in coal type, characterized in that: include: S100: Collect the current fuel feed parameters of the boiler, including online coal quality monitoring data, and obtain the coal type characteristic factor set M, which includes a feature vector composed of volatile matter, ash content, sulfur content, and lower heating value of the coal type. S200. Input the coal type characteristic factor set M into the pre-trained coal type identification model to identify the coal type L corresponding to the current burning coal type; S300. Based on the coal type L, retrieve the optimal control parameter set P that matches the coal type L from the boiler control parameter database; S400. Based on the current boiler operating status parameter R, the control parameter set P is modified to obtain the target control parameter set P' adapted to the current operating conditions; where R includes boiler load, main steam pressure, flue gas temperature and furnace negative pressure data; S500: Apply the target control parameter set P' to the boiler control and dynamically adjust the combustion process; S600 continuously monitors coal quality change trends and boiler operation feedback data. If the coal type changes or the boiler thermal deviation increases, the control strategy is re-identified and adjusted.

2. The boiler optimization control method for adaptive coal type variation according to claim 1, characterized in that: S200 includes: S201. The obtained coal type characteristic factor set M is normalized by the min-max standardization method for characteristic data of different dimensions. S202. Input the normalized feature data into the pre-trained coal type identification model. The model adopts a hybrid structure of convolutional neural network and support vector machine, wherein the convolutional layer is used to extract the deep nonlinear correlation of coal type features, and the support vector machine is used to realize multi-classification decision-making. S203. Based on the classification confidence vector output by the model, perform confidence-weighted discrimination on the recognition results. When the highest confidence is lower than the preset threshold, call the approximate feature samples for secondary matching. S204, Output the coal type L with the highest confidence level.

3. The boiler optimization control method for adaptive coal type changes according to claim 2, characterized in that: The S300 includes: S301. Receive the output coal type L and input it as a query keyword into the boiler control parameter database; S302. Match the preset control parameters in the database according to the coal type L, including the ratio of primary air to secondary air, burner tilt angle, air distribution damper opening and main combustion zone temperature setpoint. S303, Based on the current boiler operating load level The adjustable items in the matched parameter template are adaptively filtered and weighted to generate a candidate set of control parameters that are dually adapted to both coal type L and load state. ; S304, to The operating effects of the parameter combinations in the test under the same coal type and load conditions were compared and replayed, and the combination with the best thermal efficiency and stable emissions was selected as the final control parameter set P.

4. The boiler optimization control method for adaptive coal type changes according to claim 3, characterized in that: The S400 includes: S401. Collect the current operating status parameters R of the boiler, including boiler load, main steam pressure, flue gas temperature and furnace negative pressure data; S402. Construct the operating deviation vector ΔR based on parameter R, calculate the difference between the current state and the target control boundary, and obtain the dynamic deviation of each control index. S403. Input the control parameter set P and the running deviation vector ΔR into the fuzzy adaptive adjustment model; S404. Output the corrected target control parameter set P'.

5. The boiler optimization control method for adaptive coal type variation according to claim 4, characterized in that: The S500 includes: S501. Input the output target control parameter set P′ into the boiler distributed control system DCS; S502. According to the control instructions in P′, the working status of the burner, air supply system and air distribution damper and other actuators are adjusted in a coordinated manner to realize the real-time adjustment of air volume ratio, flame center position and furnace temperature.

6. The boiler optimization control method for adaptive coal type variation according to claim 1, characterized in that: The S600 includes: S601. Obtain coal intake characteristic parameters and establish a coal quality change trend model by combining time series analysis methods; S602. Synchronously collect boiler operation feedback data, including oxygen content fluctuations, furnace temperature difference and flue gas temperature deviation, and calculate the thermal deviation index T. S603. Set the coal type change judgment threshold L and the thermal deviation threshold Tt. When the coal quality trend model identifies a sudden change in coal type or the thermal deviation index T exceeds Tt, the control strategy reconstruction process is triggered. S604. Return to step S100 and re-collect the coal type characteristic factor set M.

7. An adaptive boiler optimization control system for coal type variation, used to implement the adaptive boiler optimization control method for coal type variation as described in any one of claims 1-6, characterized in that: include: The coal quality sensing module collects the current fuel feed parameters of the boiler, including online coal quality monitoring data, and obtains the coal type characteristic factor set M, which includes a feature vector composed of volatile matter, ash content, sulfur content, and lower heating value of the coal type. The coal type identification module inputs the coal type feature factor set M into the pre-trained coal type identification model to identify the coal type L corresponding to the current burning coal type. The control parameter matching module retrieves the optimal control parameter set P that matches the coal type L from the boiler control parameter database. The adaptive correction module, in conjunction with the current boiler operating status parameter R, corrects the control parameter set P to obtain the target control parameter set P' adapted to the current operating conditions; where R includes boiler load, main steam pressure, flue gas temperature and furnace negative pressure data; The boiler control execution module applies the target control parameter set P' to the boiler control and dynamically adjusts the combustion process; The system reconstructs the trigger module, continuously monitors coal quality changes and boiler operation feedback data, and re-identifies and adjusts the control strategy if the coal type changes or the boiler thermal deviation increases.

Images