Separated bunker structure design and coal blending combustion proportion dynamic regulation and control system and method

Abstract

The invention belongs to the technical field of coal-fired boiler control, and particularly relates to a system and a method for dynamically regulating and controlling the structural design of a separated coal bunker and the blending combustion proportion of fire coal. According to the method, the correction coefficient is generated through the small-scale blending combustion experiment data, and the initial proportion calculation model is calibrated, so that the proportion design conforms to the theoretical constraint and adapts to the actual combustion characteristics of the coal type, and the theoretical and actual deviation is reduced; combustion efficiency and pollutant emission trend are predicted on line based on a fire coal combustion prediction model, dynamic adjustment of a fire coal blending combustion proportion is realized in combination with a progressive adjustment rule and an extreme working condition response rule, combustion parameter fluctuation is avoided, a verified alternative scheme is called, environmental protection and combustion stability are met, response time is shortened, and the method is suitable for large-scale popularization and application. The reliability in a fault state is improved; by automatically updating the fire coal combustion prediction model, prediction-adjustment-update closed-loop optimization is realized, the regulation and control error is gradually reduced along with the operation time, and the high regulation and control precision is kept for a long time.

Description

Technical Field

[0001] This invention belongs to the field of coal-fired boiler control technology, specifically relating to a compartmentalized coal bunker structure design and a dynamic control system and method for the proportion of coal blending. Background Technology

[0002] In the industrial coal-fired power sector, coal, as a core energy source, directly impacts the overall system's operational performance through its combustion efficiency, environmental friendliness, and economic viability. With the diversification of coal resource structures and increasingly stringent environmental requirements, single-type coal combustion is gradually proving insufficient to meet demands for high efficiency, low consumption, and cleanliness. Therefore, multi-coal blending technology has become the mainstream solution. By mixing coals with different calorific values, sulfur content, and volatile matter in specific proportions, fuel costs can be balanced, pollutant emissions controlled, and combustion stability ensured.

[0003] In existing technologies, the blending of multiple coal types usually relies on fixed ratio calculation models and human experience for regulation. However, it is often based on limited laboratory data or historical operation records, pre-sets the blending ratio, and performs operations through static adjustments, which makes it difficult to fully reflect the dynamic changes in the combustion process.

[0004] Furthermore, existing technologies mainly focus on optimizing single functions, such as coal blockage treatment or pulverized coal transportation, lacking in-depth exploration of the hidden correlation patterns between blending ratios, combustion efficiency, and pollutant emissions. This results in a certain degree of imprecision and foresight in the control process. For example, Chinese Patent Publication No. CN119460463A, a coal mine anti-blockage coal bunker, focuses on solving the coal bunker blockage problem. Another example, Chinese Patent Publication No. CN109650087A, a pulverized coal storage system and control method for coal-fired power stations, mainly focuses on the automation design of pulverized coal transportation and storage processes. In summary, none of the existing technologies involve intelligent blending ratio control, thus failing to achieve full-process coverage from ratio calculation to dynamic adjustment and closed-loop optimization. Summary of the Invention

[0005] In view of this, in order to solve the problems mentioned in the background technology, a system and method for the design of a compartmentalized coal bunker structure and the dynamic control of the coal blending ratio are proposed.

[0006] The objective of this invention can be achieved through the following technical solution: The first aspect of this invention provides a compartmentalized coal bunker structure design and a dynamic control system for coal blending ratio, including a compartmentalized coal bunker structure and a dynamic control module for coal blending ratio. The compartmentalized coal bunker structure includes at least two independent compartments, a feeding unit, an anti-blocking and unblocking unit, a discharge metering unit, and a mixing and conveying unit, and is connected to the dynamic control module for coal blending ratio via a signal and works in coordination. The dynamic control module for coal blending ratio includes: a correction coefficient generation unit, an initial blending ratio determination unit, a blending ratio dynamic control unit, and a prediction model update unit.

[0007] The correction coefficient generation unit is connected to the initial blending ratio determination unit, the initial blending ratio determination unit is connected to the blending ratio dynamic control unit, and the blending ratio dynamic control unit is connected to the prediction model update unit.

[0008] The correction coefficient generation unit generates a coal blending ratio correction coefficient based on the obtained small-scale blending experimental data of several coal types to be blended.

[0009] The initial blending ratio determination unit imports the characteristic parameter data of several coal types to be blended into the initial blending ratio calculation rule model, and determines the initial blending ratio of coal by combining it with the coal blending ratio correction coefficient.

[0010] The dynamic control unit for blending ratio, based on a pre-trained coal combustion prediction model, predicts in real time the combustion efficiency and pollutant emission trends under the initial blending ratio of coal to generate forward-looking control instructions, and performs dynamic adjustment of the coal blending ratio according to preset gradual control rules and extreme operating condition response rules.

[0011] The prediction model update unit, based on the historical database, identifies the correlation patterns between coal blending ratio, combustion efficiency and pollutant emissions, and automatically updates the internal parameters of the coal combustion prediction model accordingly, thereby achieving closed-loop dynamic control of the coal ratio.

[0012] The objective of this invention can be achieved through the following technical solution: The second aspect of this invention provides a method for designing a compartmentalized coal bunker structure and dynamically controlling the coal blending ratio, including: S1, generating a coal blending ratio correction coefficient based on the obtained small-scale blending experimental data of several coal types to be blended.

[0013] S2. Import the characteristic parameter data of several coal types to be blended into the initial proportion calculation rule model, and determine the initial coal blending ratio by combining the coal blending ratio correction coefficient.

[0014] S3. Based on a pre-trained coal combustion prediction model, the combustion efficiency and pollutant emission trends under the initial coal blending ratio are predicted in real time to generate forward-looking adjustment instructions, and the coal blending ratio is dynamically adjusted according to the preset gradual adjustment rules and extreme condition response rules.

[0015] S4. Based on the historical database, identify the correlation patterns between coal blending ratio, combustion efficiency and pollutant emissions, and automatically update the internal parameters of the coal combustion prediction model accordingly to achieve closed-loop dynamic control of the coal ratio.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. The present invention generates correction coefficients through small-scale co-firing experimental data, calibrates the initial proportion calculation model, so that the proportion design not only conforms to theoretical constraints, but also adapts to the actual combustion characteristics of coal, and greatly reduces the deviation between theory and practice.

[0017] 2. This invention is based on an online prediction model of coal combustion efficiency and pollutant emission trends. It combines progressive adjustment rules and extreme condition response rules to achieve dynamic adjustment of the blending ratio, realizing active control of prediction and adjustment, avoiding fluctuations in combustion parameters, and calling alternative solutions verified by the combustion prediction model. This satisfies both environmental protection and combustion stability, and also significantly shortens the response time and improves the system reliability under fault conditions.

[0018] 3. This invention automatically updates the coal combustion prediction model by collecting and recording multi-dimensional data during the control process, realizing closed-loop optimization of prediction-adjustment-update, so that the control error gradually decreases with the running time and maintains high control accuracy in the long term. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the system module connections of the present invention.

[0021] Figure 2 This is a schematic diagram illustrating the implementation process of the method of the present invention.

[0022] Figure 3 This is a schematic diagram of the overall structure of the compartmentalized coal bunker of the present invention.

[0023] Figure 4 This is a partially enlarged view of the independent coal bunker unit of the present invention.

[0024] Figure 5 This is a schematic diagram showing the installation location of the flow sensor at the outlet of the coal feeder according to the present invention.

[0025] Figure 6 This is a schematic diagram of the structure of the mixer at the end of the coal conveying channel of the present invention. Detailed Implementation

[0026] 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.

[0027] Example 1 Please see Figure 1 As shown, this invention provides a compartmentalized coal bunker structure design and a dynamic control system for coal blending ratio, including a compartmentalized coal bunker structure and a dynamic control module for coal blending ratio. The compartmentalized coal bunker structure includes at least two independent compartments, a feeding unit, an anti-blocking and unblocking unit, a discharge metering unit, and a mixing and conveying unit. The compartmentalized coal bunker structure and the dynamic control module for coal blending ratio are connected by signals and work together. The dynamic control module for coal blending ratio includes: a correction coefficient generation unit, an initial blending ratio determination unit, a blending ratio dynamic control unit, and a prediction model update unit.

[0028] It should be further explained that the overall schematic diagram of the compartmentalized coal bunker structure is as follows: Figure 3 As shown, a partial enlarged view of the independent coal bunker unit is as follows. Figure 4 As shown in the diagram, the installation location of the flow sensor at the coal feeder outlet is as follows: Figure 5 As shown in the diagram, the structure of the mixer at the end of the coal conveying channel is as follows: Figure 6 As shown.

[0029] The correction coefficient generation unit is connected to the initial blending ratio determination unit, the initial blending ratio determination unit is connected to the blending ratio dynamic control unit, and the blending ratio dynamic control unit is connected to the prediction model update unit.

[0030] The correction coefficient generation unit generates a coal blending ratio correction coefficient based on the obtained small-scale blending experimental data of several coal types to be blended.

[0031] In a preferred feasible example of the present invention, the specific method for generating the coal blending ratio correction coefficient includes: generating the coal blending ratio correction coefficient based on the obtained small-scale blending experimental data of several coal types to be blended.

[0032] In one specific example, the small-scale co-firing experimental data includes ignition temperature characteristic curves and burnout rate characteristic curves.

[0033] It should be noted that the specific process of obtaining the small-scale co-firing experimental data of the coal to be co-firing includes: testing coal sample combinations with different co-firing ratios under different conditions using a thermogravimetric analyzer, and obtaining the ignition temperature characteristic curve and burnout rate characteristic curve corresponding to different co-firing ratios.

[0034] Ignition temperature characteristic curves and burnout rate characteristic curves of several coal types to be blended are extracted from the small-scale co-firing experimental data. These curves are then compared with the corresponding theoretical ignition temperature characteristic curves and theoretical burnout rate characteristic curves stored in the database to obtain the absolute deviation values ​​of the ignition temperature and burnout rate of several coal types to be blended under different conditions. The root mean square error of these deviations is then calculated and recorded as the coal blending ratio correction coefficient.

[0035] This invention generates correction coefficients from small-scale co-firing experimental data to calibrate the initial proportion calculation model, ensuring that the proportion design conforms to theoretical constraints and is adapted to the actual combustion characteristics of coal, thus significantly reducing the deviation between theory and practice.

[0036] The initial blending ratio determination unit imports the characteristic parameter data of several coal types to be blended into the initial blending ratio calculation rule model, and determines the initial blending ratio of coal by combining it with the coal blending ratio correction coefficient.

[0037] In a preferred feasible example of the present invention, the specific method for determining the initial coal blending ratio includes: importing the characteristic parameter data of several coal types to be blended into the initial blending ratio calculation rule model, and determining the initial coal blending ratio in combination with the coal blending ratio correction coefficient.

[0038] In one specific example, the characteristic parameter data includes industrial analysis data and elemental analysis data, wherein the industrial analysis data includes moisture, ash, volatile matter, and fixed carbon content.

[0039] The elemental analysis data includes the content of carbon, hydrogen, oxygen, nitrogen, and sulfur.

[0040] The operating parameters include furnace temperature, main steam pressure, flue gas oxygen content, and pollutant concentration.

[0041] The characteristic parameter data of several coal types to be blended are used as input into the initial proportion calculation rule model. The initial proportion calculation rule model filters the characteristic parameter data of several coal types and the target constraints after quantifying them, and outputs the preliminary proportion of several coal types. It then multiplies the preliminary proportion of several coal types by the corresponding coal blending ratio correction coefficient to obtain the initial blending ratio of several coal types, which is recorded as the initial coal blending ratio.

[0042] It should be noted that the specific method for quantifying the matching degree between the characteristic parameter data of several coal types and the target constraints is as follows: for each characteristic parameter, calculate the deviation between its actual value and the constraint threshold, such as the difference between the actual value of sulfur and the constraint threshold of 0.8%, and the difference between the actual value of volatile matter and the constraint threshold of 25%. The difference is weighted and summed to obtain the matching degree, and coal type combinations with a matching degree greater than the threshold are selected.

[0043] It should also be noted that the specific contents of the target constraint are: (1) Calorific value constraint: the theoretical calorific value of the mixed coal must match the boiler design load. For example, a boiler requires the received low calorific value of the mixed coal to be 20-25 MJ / kg.

[0044] (2) Environmental constraints: The sulfur content of the mixed coal must be ≤0.8% to meet the local emission standards and the ash content must be ≤30% to avoid slagging or excessive dust removal load.

[0045] (3) Combustion stability constraints: The volatile matter content of the mixed coal must be ≥25% to ensure ignition stability. If the volatile matter content of a single coal type does not meet the requirements, it can be achieved by blending high volatile matter coal types, such as lignite. If this cannot be achieved, a low volatile matter stable combustion plan will be triggered, such as increasing the furnace temperature.

[0046] It should be noted that the theoretical initial ratio is assumed to be... The corresponding correction factor is The corrected proportions for: .

[0047] For example, if the theoretical proportion of coal A Correction factor This means the actual ignition temperature is too high, and the proportion needs to be reduced; Coal B , If the actual burnout rate is better than the theoretical rate, the percentage can be increased, then the correction is as follows: ; .

[0048] The dynamic control unit for blending ratio, based on a pre-trained coal combustion prediction model, predicts in real time the combustion efficiency and pollutant emission trends under the initial blending ratio of coal to generate forward-looking control instructions, and performs dynamic adjustment of the coal blending ratio according to preset gradual control rules and extreme operating condition response rules.

[0049] In a preferred feasible example of the present invention, the specific training method of the pre-trained coal combustion prediction model includes: extracting coal quality analysis data, operating condition parameter data, boiler load command data, and online monitoring data of pollutant emissions from the database within the past three months as training datasets.

[0050] It should be noted that the coal quality analysis data includes characteristic parameter data and the initial coal blending ratio.

[0051] A basic model for predicting coal combustion is built using a deep learning network. The coal quality analysis data and operating parameter data are used as model inputs, and the boiler load command data and online pollutant emission monitoring data are used as model outputs. A weighted mean square error is used to set the loss function. The Adam optimizer is used, and the network weights and biases in the encoder and decoder are iteratively updated through the backpropagation algorithm until the value of the loss function converges to below a preset threshold or reaches a preset number of training cycles, thus completing the model training and obtaining the coal combustion prediction model.

[0052] For example, the preset threshold may be 0.01.

[0053] It should be noted that combustion mechanism rules are embedded in the coal combustion prediction model. For example, an increase in volatile matter leads to a decrease in ignition temperature, which in turn leads to an increase in combustion efficiency. The loss function penalizes prediction results that violate the mechanism. For example, if the prediction is that efficiency will decrease when volatile matter increases, a penalty term needs to be added. That is, when the mechanism is violated, the loss function = original error + 0.2 × absolute value of deviation.

[0054] In a preferred feasible example of the present invention, the specific method for generating the forward-looking adjustment instruction includes: taking real-time coal quality analysis data and operating parameter data for future time periods as input to the coal combustion prediction model, outputting combustion efficiency prediction curves and pollutant emission trend prediction curves under different blending ratios for future time periods, and generating blending ratio adjustment suggestions as forward-looking adjustment instructions based on these.

[0055] For example, the future time period may be within the next 2 hours.

[0056] It should be noted that the aforementioned forward-looking adjustment instructions include instructions to increase the proportion of high volatile coal types, instructions to supplement high calorific value coal types, instructions to suppress the proportion of high sulfur coal types, instructions to strengthen low nitrogen coal types, instructions to lock the proportion of stable-burning coal types, instructions to supplement coal types to prevent slagging, instructions to adjust moisture compensation, instructions to adjust hardness matching, instructions to increase calorific value under high load, instructions to adjust stable-burning under low load, instructions to replace faulty compartments, and instructions to adjust environmental protection emergency.

[0057] In a preferred feasible example of the present invention, the specific method of dynamically adjusting the coal blending ratio according to the preset progressive adjustment rules includes: setting an upper limit threshold for the single blending ratio adjustment amount, and calculating the difference between the current blending ratio and the target blending ratio when a target adjustment instruction is received.

[0058] It should be noted that the upper limit threshold is determined based on the boiler's rated load and the volatile matter characteristics of the coal, and is typically 3% to 5%. For high volatile matter coals, 3% is used to enhance stability. In this invention, the upper limit threshold can be set to 5% as an example.

[0059] If the difference is greater than the upper limit threshold, it is broken down into multiple adjustment steps. Each adjustment amount does not exceed the upper limit threshold, and the time interval between two adjacent adjustments is dynamically set according to the real-time furnace temperature fluctuation range. Thus, the target blending ratio is gradually achieved through multiple fine-tuning operations.

[0060] For example, if the real-time furnace temperature fluctuation is ≤50℃, the time interval is 30 seconds; if the real-time furnace temperature fluctuation is >50℃, the time interval is 60 seconds.

[0061] It should be noted that during the above adjustment process, combustion parameters, such as flue gas oxygen content and NOx concentration, are monitored in real time. If a certain combustion parameter fluctuates beyond the preset safety range, such as the oxygen content deviating from the target value by 1.5%, the adjustment is paused and a stable operating condition program is triggered. The program will continue to run after the parameter recovers, in order to avoid fluctuations in combustion conditions caused by continuous fine-tuning.

[0062] In a preferred feasible example of the present invention, the specific method of dynamically adjusting the coal blending ratio according to the preset extreme working condition response rules includes: real-time monitoring of the status signal of the coal feeder in the compartmentalized coal bunker, and when a coal blockage fault is detected in any coal feeder, determining whether to trigger an extreme working condition based on the duration of the fault.

[0063] As a specific example, the coal feeder status signals include, but are not limited to, abnormal current and sudden drop in speed.

[0064] The duration of the fault can be described as 10 consecutive seconds without signal recovery.

[0065] If extreme working conditions are triggered, the locking time limit is set according to the mechanical characteristics of the compartment, the compartment corresponding to the faulty coal type is automatically locked, and the residual coal clearing procedure is started.

[0066] In a specific example, the mechanical characteristics of the compartment include, but are not limited to, the response time of the pneumatic valve; the set locking time limit is typically 30 seconds to 1 minute, and can be extended to 90 seconds for large coal compartments; the residual coal clearing procedure can be to discharge residual coal through an anti-blocking and clearing unit.

[0067] The backup blending library pre-stores at least ten emergency blending schemes for different coal type combination failure scenarios to enable rapid switching.

[0068] The system retrieves and calls the alternative coal combination scheme that best matches the current boiler load from the preset backup ratio library, and inputs the alternative scheme into the coal combustion prediction model to verify its combustion efficiency and pollutant emission trend. If the prediction results meet the requirements, the system will switch; otherwise, it will call the suboptimal scheme from the backup ratio library until the requirements are met.

[0069] It should be noted that the backup blending ratio library is used to pre-store emergency coal blending ratios, which include at least ten schemes covering typical scenarios such as single-compartment coal blockage, dual-compartment collaborative failure, and failure of high-sulfur / high-volatile coal type compartments, and the schemes must meet the target constraints.

[0070] This invention uses a coal combustion prediction model to predict combustion efficiency and pollutant emission trends online. It combines progressive adjustment rules and extreme condition response rules to dynamically adjust the blending ratio, achieving proactive control through prediction and adjustment. This avoids fluctuations in combustion parameters and utilizes alternative solutions validated by the combustion prediction model. This satisfies both environmental protection and combustion stability requirements, significantly shortens response time, and improves system reliability under fault conditions.

[0071] The prediction model update unit, based on the historical database, identifies the correlation patterns between coal blending ratio, combustion efficiency and pollutant emissions, and automatically updates the internal parameters of the coal combustion prediction model accordingly, thereby achieving closed-loop dynamic control of the coal ratio.

[0072] It should be noted that the historical operation database is obtained by collecting and recording data on the characteristics of several coal types, proportioning parameters, boiler feed rate, combustion process parameters, and historical adjustment records in real time during the control process.

[0073] Specifically, the elemental analysis data and industrial analysis data of the coal collected and integrated into the furnace will be used as the coal type characteristic data; the blending ratio setting value and target coal feeding rate issued by the control unit to each coal feeder will be recorded as the blending ratio setting parameters; the actual feed amount fed back by each compartment coal feeder will be collected as the boiler feed amount data; the furnace temperature, main steam pressure, flue gas oxygen content and pollutant concentration will be collected as the combustion process parameters; and the time of each blending ratio adjustment, the values ​​before and after the adjustment, and the reason for triggering the adjustment will be recorded as the historical adjustment record.

[0074] In a preferred feasible example of the present invention, the specific method for identifying the correlation pattern between coal blending ratio, combustion efficiency and pollutant emissions includes: extracting grid load data, coal blending ratio combination data, combustion efficiency data and pollutant emission data from the historical operation database, dividing the grid load data, combustion efficiency data and pollutant emission data into several intervals, and discretizing the coal blending ratio combination data according to the proportional range.

[0075] In a specific example, the boundaries of the intervals when dividing the data into several intervals are set based on boiler design parameters and industry standards. For example, the power grid load data interval can be divided into 20% to 50%, 50% to 80%, and 80% to 100% of the rated load; the combustion efficiency data interval can be divided into ±2% error ranges; and the pollutant NOx emission data interval can be divided into 50% to 80% and 80% to 100% of the local emission standards.

[0076] The discretization is determined based on the characteristics of the coal type. For example, a 5% interval is used for high volatile matter coal types, and a 10% interval is used for low volatile matter coal types.

[0077] Using the power grid load range and coal blending ratio combination as the conditions of the association rule, and the combustion efficiency range and pollutant emission range as the results of the association rule, the support and confidence of the combustion efficiency range and pollutant emission range corresponding to various result combinations under various conditions corresponding to the power grid load range and coal blending ratio combination are calculated based on the association rule.

[0078] It should be noted that the support is the frequency of the combination in the data; the confidence is the probability that the results will occur simultaneously when the conditions are met.

[0079] If the support is less than a preset support threshold and the confidence is greater than a preset confidence threshold, the association pattern is recorded as a deviation of the calorific value balance formula for a specific coal type combination under a specific load.

[0080] For example, the deviation of the calorific value balance formula for a specific coal type combination under a specific load may be that the actual calorific value of coal type C under high load is 5% lower than the theoretical value.

[0081] If the support is greater than or equal to a preset support threshold and the confidence is greater than a preset confidence threshold, the association pattern is recorded as a new optimal matching interval.

[0082] For example, the optimal ratio range can be “load [60%, 70%] ∧ coal type A [35%, 45%] + B [55%, 65%] → efficiency [88%, 90%] ∧ NOx [80, 90] mg / m³”, where ∧ is a logical AND symbol.

[0083] In a preferred feasible example of the present invention, the specific method of automatically updating the internal parameters of the coal combustion prediction model includes: if the correlation mode shows that the calorific value balance formula of a specific coal type combination under a specific load deviates, the coal combustion prediction model automatically locates the corresponding coefficient in the formula and adjusts the coefficient value according to the deviation magnitude.

[0084] If the correlation pattern represents a new optimal ratio range, the coal combustion prediction model writes this new range into the decision rule base of the output decision layer.

[0085] It should be noted that in subsequent decision-making by the decision-making level, the coal combustion prediction model prioritizes searching for the optimal solution in the new interval, replacing the old interval, thereby improving the rationality of control.

[0086] It should be explained that by achieving closed-loop self-optimization of the control model in this way, the system's proportion control error gradually decreases as the running time increases.

[0087] This invention automatically updates the coal combustion prediction model by collecting and recording multi-dimensional data during the control process, achieving closed-loop optimization of prediction-adjustment-update, so that the control error gradually decreases over time and maintains high control accuracy over a long period of time.

[0088] Example 2 Please see Figure 2 As shown, the present invention provides a method for designing a compartmentalized coal bunker structure and dynamically controlling the coal blending ratio. The specific steps are as follows: S1. Based on the obtained small-scale blending experimental data of several coal types to be blended, a coal blending ratio correction coefficient is generated.

[0089] S2. Import the characteristic parameter data of several coal types to be blended into the initial proportion calculation rule model, and determine the initial coal blending ratio by combining the coal blending ratio correction coefficient.

[0090] S3. Based on a pre-trained coal combustion prediction model, the combustion efficiency and pollutant emission trends under the initial coal blending ratio are predicted in real time to generate forward-looking adjustment instructions, and the coal blending ratio is dynamically adjusted according to the preset gradual adjustment rules and extreme condition response rules.

[0091] S4. Based on the historical database, identify the correlation patterns between coal blending ratio, combustion efficiency and pollutant emissions, and automatically update the internal parameters of the coal combustion prediction model accordingly to achieve closed-loop dynamic control of the coal ratio.

[0092] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, and all such modifications and additions should fall within the protection scope of the present invention.

Claims

1. A compartmentalized coal bunker structure design and a dynamic control system for coal blending ratio, characterized in that: The system includes a compartmentalized coal bunker structure and a dynamic control module for coal blending ratio. The compartmentalized coal bunker structure comprises at least two independent compartments, a feeding unit, an anti-blocking and unblocking unit, a discharge metering unit, and a mixing and conveying unit. The compartmentalized coal bunker structure and the dynamic control module for coal blending ratio are connected by a signal and work together. The dynamic control module for coal blending ratio includes: The correction coefficient generation unit generates a coal blending ratio correction coefficient based on the obtained small-scale blending experimental data of several coal types to be blended. The initial blending ratio determination unit imports the characteristic parameter data of several coal types to be blended into the initial blending ratio calculation rule model, and determines the initial blending ratio of coal by combining it with the coal blending ratio correction coefficient. The dynamic control unit for blending ratio, based on a pre-trained coal combustion prediction model, predicts the combustion efficiency and pollutant emission trends under the initial blending ratio of coal in real time to generate forward-looking control instructions, and performs dynamic adjustment of the coal blending ratio according to the preset gradual control rules and extreme condition response rules. The prediction model update unit, based on the historical database, identifies the correlation patterns between coal blending ratio, combustion efficiency and pollutant emissions, and automatically updates the internal parameters of the coal combustion prediction model accordingly, thereby achieving closed-loop dynamic control of the coal ratio.

2. The compartmentalized coal bunker structure design and dynamic control system for coal blending ratio according to claim 1, characterized in that: The specific methods for generating the coal blending ratio correction coefficient include: Based on the small-scale co-firing experimental data of several coal types to be co-firing, a coal co-firing ratio correction coefficient is generated. Ignition temperature characteristic curves and burnout rate characteristic curves of several coal types to be blended are extracted from the small-scale co-firing experimental data. These curves are then compared with the corresponding theoretical ignition temperature characteristic curves and theoretical burnout rate characteristic curves stored in the database to obtain the absolute deviation values ​​of the ignition temperature and burnout rate of several coal types to be blended under different conditions. The root mean square error of these deviations is then calculated and recorded as the coal blending ratio correction coefficient.

3. The compartmentalized coal bunker structure design and dynamic control system for coal blending ratio according to claim 1, characterized in that: The specific methods for determining the initial coal blending ratio include: The characteristic parameter data of several coal types to be blended are imported into the initial proportion calculation rule model, and the initial coal blending ratio is determined by combining the coal blending ratio correction coefficient. The characteristic parameter data of several coal types to be blended are used as input into the initial proportion calculation rule model. The initial proportion calculation rule model filters the characteristic parameter data of several coal types and the target constraints after quantifying them, and outputs the preliminary proportion of several coal types. It then multiplies the preliminary proportion of several coal types by the corresponding coal blending ratio correction coefficient to obtain the initial blending ratio of several coal types, which is recorded as the initial coal blending ratio.

4. The compartmentalized coal bunker structure design and dynamic control system for coal blending ratio according to claim 1, characterized in that: The specific training methods for the pre-trained coal combustion prediction model include: The training dataset was extracted from the database, which contained coal quality analysis data, operating parameter data, boiler load command data, and online pollutant emission monitoring data from the past three months. A basic model for predicting coal combustion is built using a deep learning network. The coal quality analysis data and operating parameter data are used as model inputs, and the boiler load command data and online pollutant emission monitoring data are used as model outputs. A weighted mean square error is used to set the loss function. The Adam optimizer is used, and the network weights and biases in the encoder and decoder are iteratively updated through the backpropagation algorithm until the value of the loss function converges to below a preset threshold or reaches a preset number of training cycles, thus completing the model training and obtaining the coal combustion prediction model.

5. The compartmentalized coal bunker structure design and dynamic control system for coal blending ratio according to claim 1, characterized in that: The specific methods for generating forward-looking adjustment instructions include: Real-time coal quality analysis data and operating parameter data for future time periods are used as inputs to the coal combustion prediction model. The model outputs combustion efficiency prediction curves and pollutant emission trend prediction curves under different blending ratios for future time periods, and generates blending ratio adjustment suggestions as forward-looking adjustment instructions.

6. The compartmentalized coal bunker structure design and dynamic control system for coal blending ratio according to claim 1, characterized in that: The specific methods for dynamically adjusting the coal blending ratio according to the preset gradual adjustment rules include: Set an upper limit threshold for the single blending ratio adjustment. When a target adjustment command is received, calculate the difference between the current blending ratio and the target blending ratio. If the difference is greater than the upper limit threshold, it is broken down into multiple adjustment steps. Each adjustment amount does not exceed the upper limit threshold, and the time interval between two adjacent adjustments is dynamically set according to the real-time furnace temperature fluctuation range. Thus, the target blending ratio is gradually achieved through multiple fine-tuning operations.

7. The compartmentalized coal bunker structure design and dynamic control system for coal blending ratio according to claim 1, characterized in that: The specific methods for dynamically adjusting the coal blending ratio based on preset extreme operating condition response rules include: Real-time monitoring of the status signals of the coal feeders in the compartmentalized coal bunker; when a coal blockage fault signal is detected in any coal feeder, the system determines whether to trigger extreme working conditions based on the duration of the fault. If extreme working conditions are triggered, the locking time limit is set according to the mechanical characteristics of the compartment, the compartment corresponding to the faulty coal type is automatically locked, and the residual coal clearing procedure is started. The backup blending library pre-stores at least ten emergency blending schemes for different coal type combination failure scenarios to enable rapid switching; The system retrieves and calls the alternative coal combination scheme that best matches the current boiler load from the preset backup ratio library, and inputs the alternative scheme into the coal combustion prediction model to verify its combustion efficiency and pollutant emission trend. If the prediction results meet the requirements, the system will switch; otherwise, it will call the suboptimal scheme from the backup ratio library until the requirements are met.

8. The compartmentalized coal bunker structure design and dynamic control system for coal blending ratio according to claim 1, characterized in that: The specific methods for identifying the correlation patterns between coal blending ratio, combustion efficiency, and pollutant emissions include: Extract grid load data, coal blending ratio data, combustion efficiency data, and pollutant emission data from the historical operation database. Divide the grid load data, combustion efficiency data, and pollutant emission data into several intervals and discretize the coal blending ratio data according to the proportional range. The grid load range and coal blending ratio combination are used as the conditions of the association rule, and the combustion efficiency range and pollutant emission range are used as the results of the association rule. Based on the association rule, the support and confidence of the combustion efficiency range and pollutant emission range corresponding to multiple result combinations under the various combination conditions of grid load range and coal blending ratio combination are calculated. If the support is less than a preset support threshold and the confidence is greater than a preset confidence threshold, the association pattern is recorded as a deviation of the calorific value balance formula for a specific coal type combination under a specific load. If the support is greater than or equal to a preset support threshold and the confidence is greater than a preset confidence threshold, the association pattern is recorded as a new optimal matching interval.

9. The compartmentalized coal bunker structure design and dynamic control system for coal blending ratio according to claim 8, characterized in that: The specific methods for automatically updating the internal parameters of the coal combustion prediction model include: If the correlation pattern shows that the calorific value balance formula of a specific coal type combination under a specific load deviates, the coal combustion prediction model will automatically locate the corresponding coefficient in the formula and adjust the coefficient value according to the deviation magnitude. If the correlation pattern represents a new optimal ratio range, the coal combustion prediction model writes this new range into the decision rule base of the output decision layer.

10. A method for designing a compartmentalized coal bunker structure and dynamically controlling the coal blending ratio, characterized in that: include: S1. Based on the small-scale co-firing experimental data of several coal types to be co-firing, generate the coal co-firing ratio correction coefficient; S2. Import the characteristic parameter data of several coal types to be blended into the initial proportion calculation rule model, and determine the initial coal blending ratio in combination with the coal blending ratio correction coefficient. S3. Based on a pre-trained coal combustion prediction model, the combustion efficiency and pollutant emission trends under the initial coal blending ratio are predicted in real time to generate forward-looking adjustment instructions, and the coal blending ratio is dynamically adjusted according to the preset gradual adjustment rules and extreme condition response rules. S4. Based on the historical database, identify the correlation patterns between coal blending ratio, combustion efficiency and pollutant emissions, and automatically update the internal parameters of the coal combustion prediction model accordingly to achieve closed-loop dynamic control of the coal ratio.

Images