Circulating fluidized bed boiler combustion self-optimization control method and system

Abstract

This invention relates to the field of boiler combustion optimization control technology, specifically to an autonomous optimization control method and system for circulating fluidized bed (CFB) boiler combustion. Addressing the problems of existing CFB boilers, such as easy interference between various adjustment actions, untimely switching of the main control objective, and insufficient identification of operational boundary risks under load fluctuations, coal quality changes, and changes in return material status, this invention collects operating parameters for the current control cycle, determines the main control objective, generates a set of discrete adjustment actions, and filters them based on a boiler combustion control constraint template to obtain a set of permissible adjustment actions. Then, it performs conflict determination on any two discrete adjustment actions, constructs an undirected graph of action conflicts, extracts a set of candidate adjustment action combinations through independent set extraction, and determines the target adjustment action combination by combining the improvement amount of the main control objective and the change in the boiler's operational safety margin, and executes the adjustment. This invention is used for autonomous optimization control of the combustion process in a CFB boiler.

Description

Technical Field

[0001] This invention relates to the field of boiler combustion optimization control technology, specifically to a method and system for autonomous optimization control of combustion in circulating fluidized bed boilers. Background Technology

[0002] Circulating fluidized bed (CFB) boilers, due to their strong fuel adaptability, high combustion efficiency, and relatively low pollutant emissions, have been widely used in power generation, heating, chemical industry, and resource recycling. With the increase in boiler capacity and the increasing complexity of operating conditions, combustion control methods have evolved from early reliance on manual experience-based adjustments to automatic control based on instrument monitoring, and further to coordinated control based on DCS and optimized control aimed at energy conservation and emission reduction. Currently, combustion control of CFB boilers has become a crucial technical aspect for the safe, stable, and economical operation of boilers.

[0003] Existing circulating fluidized bed boiler combustion control methods mostly employ fixed rules, single-variable loop regulation, or experience-based coordinated control. Under conditions such as load fluctuations, coal quality changes, and changes in return material status, mutual interference between coal feeding, primary air, secondary air, and return material regulation is prone to occur. It is difficult to coordinate and select and optimize multiple discrete regulation actions, resulting in untimely switching of the main control target, insufficient identification of operational boundary risks, and difficulty in quickly forming a stable and effective regulation combination when combustion organization is unbalanced. Consequently, the safety, stability, and economy of boiler operation are affected. Therefore, a circulating fluidized bed boiler combustion autonomous optimization control method and system are needed to solve the above problems. Summary of the Invention

[0004] (a) Technical problems to be solved To address the shortcomings of existing technologies, this invention provides a method and system for autonomous optimization control of combustion in circulating fluidized bed boilers, thus solving the aforementioned problems.

[0005] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: a method for autonomous optimization control of combustion in a circulating fluidized bed boiler, comprising: S1. Collect the operating parameters of the circulating fluidized bed boiler during the current control cycle; S2. Determine the main control target for the current control cycle based on the operating parameters; S3. Based on the operating parameters and the main control target, generate a set of corresponding discrete adjustment actions for the boiler combustion adjustment execution amount; S4. Substitute the operating parameters into the pre-established boiler combustion control constraint template, perform constraint screening on the discrete adjustment action set, and obtain the allowable adjustment action set; S5. Perform conflict determination on any two discrete adjustment actions in the set of allowed adjustment actions, and construct an undirected graph of action conflicts based on the conflict determination results; S6. Extract independent sets from the undirected graph of action conflict to obtain a set of candidate adjustment action combinations; S7. Based on the improvement amount of each candidate adjustment action combination in the candidate adjustment action combination set to the main control target and the change amount to the boiler operation safety margin, determine the target adjustment action combination, and control the circulating fluidized bed boiler to perform adjustment according to the target adjustment action combination.

[0006] Furthermore, the operating parameters include bed temperature, bed pressure difference, furnace outlet oxygen content, main steam pressure, coal feed rate, primary air volume, secondary air volume, and return valve opening.

[0007] Furthermore, S2 includes: When the bed pressure difference is less than the first warning threshold, restoring the bed pressure difference will be determined as the main control target; When the bed pressure difference is not less than the first warning threshold and the bed temperature is greater than the second warning threshold, reducing the bed temperature will be determined as the main control target. When the bed pressure difference is not less than the first warning threshold, the bed temperature is not greater than the second warning threshold, and the main steam pressure deviation is greater than the third threshold, the corrected main steam pressure will be determined as the main control target. If none of the aforementioned conditions are met, the oxygen content at the furnace outlet will be determined as the primary control target.

[0008] Furthermore, the boiler combustion regulation parameters include coal feed rate, primary air volume, secondary air volume, and return valve opening; S3 includes: The system generates actions to increase and decrease coal feed based on the coal feed rate, actions to increase and decrease primary air volume based on the primary air volume, actions to increase and decrease secondary air volume based on the secondary air volume, and actions to increase and decrease the return valve opening based on the return valve opening degree.

[0009] Furthermore, the boiler combustion control constraint template includes bed temperature constraint, bed pressure difference constraint, furnace outlet oxygen constraint, main steam pressure constraint, coal feed variation constraint, air-coal matching constraint, staged air distribution constraint, and return material coordination constraint.

[0010] Furthermore, the conflict determination result includes the determination result corresponding to the physical mutual exclusion relationship; The physical mutual exclusion relationship includes the relationship that the increase action and the decrease action corresponding to the same boiler combustion regulation execution amount cannot be realized at the same time within the same control cycle.

[0011] Furthermore, the conflict determination result includes the determination result corresponding to the running boundary approximation relationship; The boiler operation safety margin is determined based on the first boundary distance between the bed temperature and the preset upper limit boundary, the second boundary distance between the bed pressure difference and the preset lower limit boundary, the third boundary distance between the furnace outlet oxygen quantity and the preset target interval boundary, and the fourth boundary distance between the main steam pressure and the preset lower limit boundary in the one-step prediction results corresponding to any two discrete adjustment actions. The operational boundary approximation relationship is determined based on the first boundary distance and the second boundary distance; When the first boundary distance is less than the first threshold and the second boundary distance is less than the second threshold, it is determined that there is a running boundary approximation relationship between any two discrete adjustment actions.

[0012] Furthermore, the conflict determination result includes the determination result corresponding to the imbalance relationship of combustion organization; The combustion organization imbalance includes the imbalance between air and coal, the imbalance between primary and secondary air distribution, and the imbalance between return material and coal feeding. The air-coal matching imbalance index corresponding to the air-coal matching imbalance relationship, the graded air distribution imbalance index corresponding to the primary air and secondary air graded air distribution imbalance relationship, and the return material coordination imbalance index corresponding to the return material and coal feeding coordination imbalance relationship are determined based on the one-step prediction results corresponding to any two discrete adjustment actions. The preset allowable range is determined by taking the intersection of the allowable range given in the boiler operating procedure, the historical steady-state operating condition statistical range, and the boundary range of the mechanism model; When any of the following imbalance indicators exceeds the preset allowable range: the air-coal matching imbalance indicator, the graded air distribution imbalance indicator, and the return material coordination imbalance indicator, it is determined that there is a combustion organization imbalance relationship between any two discrete adjustment actions.

[0013] Furthermore, S6 includes: Assign node weights to each node in the undirected graph of action conflict, extract multiple independent sets based on the node weights, sort them according to the total node weights corresponding to each independent set, and determine the candidate set of adjustment action combinations. The node weights are determined based on the improvement of the main control objective by the corresponding discrete adjustment action and the change in the boiler operation safety margin.

[0014] This invention also provides an autonomous combustion optimization control system for circulating fluidized bed boilers, comprising: The operating parameter acquisition module is used to collect the operating parameters of the circulating fluidized bed boiler during the current control cycle. The main control target determination module is used to determine the main control target for the current control cycle based on the operating parameters. The discrete action generation module is used to generate a set of corresponding discrete adjustment actions for the boiler combustion adjustment execution amount based on the operating parameters and the main control target. The constraint filtering module is used to substitute the operating parameters into a pre-established boiler combustion control constraint template, perform constraint filtering on the discrete adjustment action set, and obtain the set of allowed adjustment actions. The conflict graph construction module is used to determine the conflict between any two discrete adjustment actions in the set of allowed adjustment actions, and to construct an undirected graph of action conflicts based on the conflict determination results. The candidate combination extraction module is used to extract independent sets from the action conflict undirected graph to obtain a set of candidate adjustment action combinations. The target execution module is used to determine the target adjustment action combination based on the improvement amount of each candidate adjustment action combination in the candidate adjustment action combination set to the main control target and the change amount to the boiler operation safety margin, and to control the circulating fluidized bed boiler to perform adjustment according to the target adjustment action combination.

[0015] (III) Beneficial Effects Compared with the prior art, the present invention provides a method and system for autonomous optimization control of combustion in circulating fluidized bed boilers, which has the following beneficial effects: 1. The circulating fluidized bed boiler combustion autonomous optimization control method and system generates a discrete set of adjustment actions based on the operating parameters and main control objectives of the current control cycle. The operating parameters are then substituted into a pre-established boiler combustion control constraint template to perform constraint screening on the discrete set of adjustment actions. Under conditions of load fluctuation, coal quality change, and return material status change, adjustment actions that do not meet the operating constraints can be eliminated in advance. This avoids the problem of unsuitable actions participating in the adjustment simultaneously in traditional fixed rule or experience-based coordinated control, thereby improving the targeting of action generation after the main control objective is switched and enhancing the adaptability of the combustion adjustment process to complex operating conditions.

[0016] 2. The circulating fluidized bed boiler combustion autonomous optimization control method and system can screen out compatible candidate adjustment action combinations from multiple candidate adjustment actions by performing conflict determination on any two discrete adjustment actions in the set of allowed adjustment actions, constructing an action conflict undirected graph based on the conflict determination results, and then extracting independent sets from the action conflict undirected graph. This reduces the mutual interference between coal feeding, primary air, secondary air and return material adjustment, and solves the problems in the existing technology of difficulty in coordinating and optimizing multiple adjustment actions and difficulty in quickly forming a stable and effective adjustment combination when the combustion organization is unbalanced. This improves the coordination and stability of boiler combustion control.

[0017] 3. The circulating fluidized bed boiler combustion autonomous optimization control method and system determines the target adjustment action combination based on the improvement of each candidate adjustment action combination on the main control target and the change in the boiler operation safety margin. It can take into account the operational boundary risks while pursuing rapid correction of the main control target, avoiding the problems of untimely switching of the main control target, insufficient identification of operational boundary risks, and inconsistencies in the adjustment process in traditional control methods. This is conducive to improving the safety, stability and economy of circulating fluidized bed boiler operation. Attached Figure Description

[0018] Figure 1 A schematic diagram illustrating the steps of the autonomous optimization control method for combustion in a circulating fluidized bed boiler provided by this invention.

[0019] Figure 2 A schematic flowchart of the autonomous optimization control method for combustion in a circulating fluidized bed boiler provided by the present invention.

[0020] Figure 3 This is a schematic diagram of the autonomous combustion optimization control system for circulating fluidized bed boilers provided by the present invention. Detailed Implementation

[0021] 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.

[0022] To enable those skilled in the art to better understand the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0023] Please see Figure 1-2 , Figure 1 A schematic diagram illustrating the steps of the autonomous optimization control method for combustion in a circulating fluidized bed boiler provided by the present invention; Figure 2 This invention provides a flowchart illustrating the autonomous optimization control method for combustion in a circulating fluidized bed boiler. The invention includes: S1. Collect the operating parameters of the circulating fluidized bed boiler during the current control cycle; Specifically, at the start of the current control cycle, operating data is synchronously acquired from the boiler measuring points and the actuator feedback channels, and time alignment is completed to form the operating parameter set corresponding to the current control cycle.

[0024] Furthermore, in one embodiment of the present invention, the operating parameters include bed temperature, bed pressure difference, furnace outlet oxygen content, main steam pressure, coal feed rate, primary air volume, secondary air volume, and return valve opening.

[0025] Specifically, bed temperature represents the thermal state of bed combustion, bed pressure difference represents the fluidization state of bed material, furnace outlet oxygen content represents the oxygen content level after combustion, main steam pressure represents the steam side output state; coal feed rate represents the fuel input level, primary air volume and secondary air volume represent the air distribution state, and return valve opening degree represents the return material adjustment state.

[0026] S2. Determine the main control target for the current control cycle based on the operating parameters; Specifically, within each control cycle, the controller does not process multiple control objectives in parallel. Instead, it first classifies and classifies the operating parameters, and then selects the single primary control objective for the current cycle. The classification order is set as bed pressure difference, bed temperature, main steam pressure, and furnace outlet oxygen content. Bed pressure difference directly reflects the fluidization state of the bed material. When this value enters the abnormal zone first, if subsequent coal feeding and air distribution adjustments continue to focus on other objectives, the fluidization state is likely to deviate further. Bed temperature reflects the heat release level of the bed. When the pressure difference has not entered the priority processing zone but the bed temperature has exceeded the control limit, the bed thermal state needs to be addressed first. Main steam pressure reflects the steam-side output deviation. When the fluidization state and bed thermal state have not triggered priority control, steam-side correction is then applied. Furnace outlet oxygen content is used to implement combustion air-side correction when none of the aforementioned primary states have exceeded the limits. By determining the primary control objective in this order, the direction of action generation within each control cycle remains singular, avoiding the mutual restraint of actions caused by simultaneous parallel adjustments around fluidization, thermal state, and steam output within the same cycle.

[0027] Furthermore, in one embodiment provided by the present invention, S2 includes: When the bed pressure difference is less than the first warning threshold, restoring the bed pressure difference will be determined as the main control target; When the bed pressure difference is not less than the first warning threshold and the bed temperature is greater than the second warning threshold, reducing the bed temperature will be determined as the main control target. When the bed pressure difference is not less than the first warning threshold, the bed temperature is not greater than the second warning threshold, and the main steam pressure deviation is greater than the third threshold, the corrected main steam pressure will be determined as the main control target. If none of the aforementioned conditions are met, the oxygen content at the furnace outlet will be determined as the primary control target.

[0028] Specifically, the first warning threshold is determined based on the lower limit of the bed pressure difference under design conditions and the historical steady-state pressure difference range; the second warning threshold is determined based on the upper limit of the bed temperature under design conditions and the historical steady-state bed temperature range; the third threshold is determined based on the allowable fluctuation range of the main steam pressure and the historical steady-state pressure deviation range, wherein the main steam pressure deviation is the absolute value of the deviation of the current main steam pressure relative to the preset target pressure. When the bed pressure difference is lower than the first warning threshold, fluidization is restored first; when the bed pressure difference is not triggered but the bed temperature is higher than the second warning threshold, the bed temperature is reduced first; when neither of the first two is triggered but the main steam pressure deviation is greater than the third threshold, the main steam pressure is corrected first; in other cases, the oxygen content at the furnace outlet is taken as the main control target.

[0029] S3. Based on the operating parameters and the main control target, generate a set of corresponding discrete adjustment actions for the boiler combustion adjustment execution amount; Specifically, the controller generates an initial discrete set of adjustment actions based on the main control objective of the current control cycle, focusing on the boiler combustion regulation execution quantities. Each execution quantity is then discretized into a finite number of increase and decrease actions. This initial discrete set of adjustment actions includes at least increase and decrease actions corresponding to coal feed rate, primary air volume, secondary air volume, and return valve opening, respectively. Subsequently, the initial discrete set of adjustment actions is constrained and filtered using current operating parameters and the boiler combustion control constraint template to obtain the set of permissible adjustment actions participating in the current cycle's regulation. After discretization, subsequent constraint filtering, conflict determination, and combination extraction are all performed based on a unified action granularity.

[0030] Furthermore, in one embodiment of the present invention, the boiler combustion regulation execution quantity includes coal feed rate, primary air volume, secondary air volume, and return valve opening degree; S3 includes: The system generates actions to increase and decrease coal feed based on the coal feed rate, actions to increase and decrease primary air volume based on the primary air volume, actions to increase and decrease secondary air volume based on the secondary air volume, and actions to increase and decrease the return valve opening based on the return valve opening degree.

[0031] Specifically, the coal feed rate, primary air volume, secondary air volume, and return valve opening correspond to fuel input, bottom fluidized air supply, upper tiered air distribution, and return material regulation, respectively. For each execution quantity, an increase action and a decrease action are set, ensuring that each execution quantity corresponds to a set of discrete regulation actions with clear directions within the current control cycle. The resulting set of discrete regulation actions retains the main control direction of boiler combustion regulation and facilitates subsequent conflict determination between any two discrete regulation actions.

[0032] S4. Substitute the operating parameters into the pre-established boiler combustion control constraint template, perform constraint screening on the discrete adjustment action set, and obtain the allowable adjustment action set; Specifically, after the set of discrete control actions is generated, the operating parameters of the current control cycle are input into the boiler combustion control constraint template, and each discrete control action is checked to see if it meets the operating limits under the current conditions. Discrete control actions that meet the constraints are retained, while those that do not meet the constraints are removed, resulting in the set of permissible control actions.

[0033] Furthermore, in one embodiment of the present invention, the boiler combustion control constraint template includes bed temperature constraint, bed pressure difference constraint, furnace outlet oxygen constraint, main steam pressure constraint, coal feed variation constraint, air-coal matching constraint, staged air distribution constraint, and return material coordination constraint.

[0034] Specifically, for each discrete control action in the set of discrete control actions, the current operating parameters and the discrete control action are first input into a pre-established one-step prediction model to obtain the predicted bed temperature, predicted bed pressure difference, predicted furnace outlet oxygen, predicted main steam pressure, predicted coal feed rate, predicted primary air volume, predicted secondary air volume, and predicted return valve opening for the discrete control action in the next control step; then, each prediction result is compared with the corresponding constraint item in the boiler combustion control constraint template. If the predicted bed temperature is higher than the preset upper limit, or the predicted bed pressure difference is lower than the preset lower limit, or the predicted furnace outlet oxygen exceeds the preset target range, or the predicted main steam pressure enters the preset abnormal fluctuation range, then the discrete adjustment action is determined to not meet the state constraints. If the coal feed change corresponding to the discrete adjustment action exceeds the allowable change range for a single cycle, or the air-coal matching deviation between the predicted coal feed and the predicted primary air volume exceeds the allowable deviation, or the staged air distribution ratio deviation between the predicted primary air volume and the predicted secondary air volume exceeds the allowable deviation, or the coordination deviation between the predicted return valve opening and the predicted coal feed exceeds the allowable deviation, then the discrete adjustment action is determined to not meet the coordination constraints. Discrete adjustment actions that meet all constraints are retained, and discrete adjustment actions that do not meet any constraint are eliminated, thus obtaining the set of allowable adjustment actions.

[0035] S5. Perform conflict determination on any two discrete adjustment actions in the set of allowed adjustment actions, and construct an undirected graph of action conflicts based on the conflict determination results; Specifically, taking the set of allowed adjustment actions as the objects to be judged, two discrete adjustment actions are randomly selected to form an action pair. Each pair is then judged to see if it can be simultaneously determined within the same control cycle. If they cannot be simultaneously determined, the action pair is judged as a conflicting action pair, and the two discrete adjustment actions are mapped to two nodes in the action conflict undirected graph, with an edge established between the two nodes. If they can be simultaneously determined, no edge is established. After all action pairs are judged, the action conflict undirected graph is obtained. Thus, nodes in the action conflict undirected graph represent discrete adjustment actions, and edges represent conflict relationships between two discrete adjustment actions.

[0036] Furthermore, in one embodiment provided by the present invention, the conflict determination result includes the determination result corresponding to the physical mutual exclusion relationship; The physical mutual exclusion relationship includes the relationship that the increase action and the decrease action corresponding to the same boiler combustion regulation execution amount cannot be realized at the same time within the same control cycle.

[0037] Specifically, for the same boiler combustion regulation execution quantity, only one regulation direction is retained. When the increase and decrease actions corresponding to the same execution quantity occur simultaneously, they will form opposite regulation directions within the same control cycle. Therefore, they are directly determined to be physically mutually exclusive, and edges are established for the nodes corresponding to these two discrete regulation actions in the action conflict undirected graph. This determination does not depend on subsequent predicted quantities and is completed directly based on the action definition.

[0038] Furthermore, in one embodiment provided by the present invention, the conflict determination result includes the determination result corresponding to the running boundary approximation relationship; The boiler operation safety margin is determined based on the first boundary distance between the bed temperature and the preset upper limit boundary, the second boundary distance between the bed pressure difference and the preset lower limit boundary, the third boundary distance between the furnace outlet oxygen quantity and the preset target interval boundary, and the fourth boundary distance between the main steam pressure and the preset lower limit boundary in the one-step prediction results corresponding to any two discrete adjustment actions. The operational boundary approximation relationship is determined based on the first boundary distance and the second boundary distance; When the first boundary distance is less than the first threshold and the second boundary distance is less than the second threshold, it is determined that there is a running boundary approximation relationship between any two discrete adjustment actions.

[0039] Specifically, the one-step prediction result is used to characterize the state changes that any two discrete adjustment actions may cause in the next control step. The input of the one-step prediction model includes the bed temperature, bed pressure difference, furnace outlet oxygen content, main steam pressure, coal feed rate, primary air volume, secondary air volume, return valve opening degree, and the action codes corresponding to any two discrete adjustment actions in the current control cycle. The output includes the predicted bed temperature, predicted bed pressure difference, predicted furnace outlet oxygen content, predicted main steam pressure, predicted coal feed rate, predicted primary air volume, predicted secondary air volume, and predicted return valve opening degree for the next control step. In one embodiment, the action code is used to characterize the execution quantity category and adjustment direction corresponding to the discrete adjustment action; for coal feed increase action, coal feed decrease action, primary air increase action, primary air decrease action, secondary air increase action, secondary air decrease action, return valve opening increase action, and return valve opening decrease action, they can be encoded using a unique thermal coding method or an execution quantity identifier plus direction identifier method, respectively. The one-step prediction model can be established using a regression model trained on historical operating data, or it can be established using a hybrid prediction model that combines boiler mechanism equations with historical data correction.

[0040] The first boundary distance represents the remaining distance from the predicted bed temperature to the upper limit boundary of the bed temperature; the second boundary distance represents the remaining distance from the predicted bed pressure difference to the lower limit boundary of the bed pressure difference; the third boundary distance represents the remaining distance from the predicted furnace outlet oxygen quantity to the nearest boundary of the target interval; and the fourth boundary distance represents the remaining distance from the predicted main steam pressure to the lower limit boundary of the main steam pressure. To eliminate the influence of different parameter dimensions, the first, second, third, and fourth boundary distances are divided by the preset safety bandwidth of the corresponding parameters to obtain the first, second, third, and fourth normalized boundary distances. The preset safety bandwidth is determined based on the difference between the control boundary and the warning boundary of the corresponding operating parameter, or jointly determined based on the allowable fluctuation range of the design operating conditions, the historical steady-state fluctuation range, and the warning margin in the operating procedures. The minimum value among the first, second, third, and fourth normalized boundary distances is determined as the boiler operating safety margin, used to characterize the safety level closest to the risk boundary in the one-step prediction result corresponding to any two discrete adjustment actions.

[0041] The operational boundary approximation relationship is determined based on the first boundary distance and the second boundary distance, and is used to identify whether an action simultaneously pushes the bed's thermal state and fluidization state close to the risk boundary. The first threshold is determined based on the thermal safety margin reserved outside the upper limit boundary of the bed temperature, and the second threshold is determined based on the fluidization safety margin reserved outside the lower limit boundary of the bed pressure difference. When the first boundary distance is less than the first threshold and the second boundary distance is less than the second threshold, it is determined that there is an operational boundary approximation relationship between any two discrete adjustment actions, and an edge is established for the corresponding two nodes in the action conflict undirected graph.

[0042] Furthermore, in one embodiment provided by the present invention, the conflict determination result includes the determination result corresponding to the imbalance relationship of combustion organization; The combustion organization imbalance includes the imbalance between air and coal, the imbalance between primary and secondary air distribution, and the imbalance between return material and coal feeding. The air-coal matching imbalance index corresponding to the air-coal matching imbalance relationship, the graded air distribution imbalance index corresponding to the primary air and secondary air graded air distribution imbalance relationship, and the return material coordination imbalance index corresponding to the return material and coal feeding coordination imbalance relationship are determined based on the one-step prediction results corresponding to any two discrete adjustment actions. The preset allowable range is determined by taking the intersection of the allowable range given in the boiler operating procedure, the historical steady-state operating condition statistical range, and the boundary range of the mechanism model; When any of the following imbalance indicators exceeds the preset allowable range: the air-coal matching imbalance indicator, the graded air distribution imbalance indicator, and the return material coordination imbalance indicator, it is determined that there is a combustion organization imbalance relationship between any two discrete adjustment actions.

[0043] Specifically, the air-coal matching imbalance index is used to characterize the degree of mismatch between fuel input and primary air supply. It can be calculated based on the predicted coal feed rate and predicted primary air volume from the first-step prediction result. In one embodiment, a target matching function between the coal feed rate and primary air volume is first fitted based on historical steady-state operating conditions to obtain the target primary air volume corresponding to the predicted coal feed rate. Then, the absolute value of the difference between the predicted primary air volume and the target primary air volume is determined as the air-coal matching imbalance index.

[0044] The graded air distribution imbalance index is used to characterize the degree of deviation between the primary and secondary air distribution ratios. It can be calculated based on the predicted primary air volume and the predicted secondary air volume from the one-step prediction result. In one embodiment, the ratio of the predicted secondary air volume to the predicted primary air volume is used as the predicted graded air distribution ratio, and the absolute value of the difference between the predicted graded air distribution ratio and the preset target graded air distribution ratio is determined as the graded air distribution imbalance index.

[0045] The return material coordination imbalance index is used to characterize the degree of coordination deviation between return material regulation and coal feeding regulation. It can be calculated based on the predicted return material valve opening and the predicted coal feeding rate from the one-step prediction result. In one embodiment, a target coordination function between the coal feeding rate and the return material valve opening is first established based on historical steady-state operating conditions or mechanism relationships to obtain the target return material valve opening corresponding to the predicted coal feeding rate. Then, the absolute value of the difference between the predicted return material valve opening and the target return material valve opening is determined as the return material coordination imbalance index.

[0046] The preset allowable intervals include preset allowable intervals corresponding to the air-coal matching imbalance index, the staged air distribution imbalance index, and the return material coordination imbalance index, respectively. Each preset allowable interval is determined by taking the intersection of the allowable interval given in the boiler operation procedure, the historical steady-state operating condition statistical interval, and the boundary interval of the mechanism model. If any of the air-coal matching imbalance index, staged air distribution imbalance index, or return material coordination imbalance index exceeds its corresponding preset allowable interval, it is determined that there is a combustion organization imbalance relationship between any two discrete adjustment actions, and an edge is established for the corresponding two nodes in the action conflict undirected graph.

[0047] S6. Extract independent sets from the undirected graph of action conflict to obtain a set of candidate adjustment action combinations; Specifically, in an undirected graph of action conflicts, nodes represent discrete adjustment actions, and edges represent two discrete adjustment actions that cannot be simultaneously executed within the same control cycle. During independent set extraction, node combinations where no two nodes are connected by an edge are selected as action combinations that can be executed simultaneously. The resulting node combinations do not contain conflicting actions and can be used as candidate adjustment action combinations.

[0048] Furthermore, in one embodiment provided by the present invention, S6 includes: Assign node weights to each node in the undirected graph of action conflict, extract multiple independent sets based on the node weights, sort them according to the total node weights corresponding to each independent set, and determine the candidate set of adjustment action combinations. The node weights are determined based on the improvement of the main control objective by the corresponding discrete adjustment action and the change in the boiler operation safety margin.

[0049] Specifically, node weights are used to characterize the priority of individual discrete control actions. For any node in the undirected graph of action conflicts, the improvement value of the main control objective is first obtained based on the predicted improvement of the main control objective when the discrete control action corresponding to that node is executed alone; then, the change in safety margin is obtained based on the difference between the predicted boiler operating safety margin corresponding to the discrete control action corresponding to that node and the boiler operating safety margin in the current control cycle. In one embodiment, the node weights are determined according to the following rules: the node weight equals the first weight coefficient multiplied by the improvement value of the main control objective, minus the second weight coefficient multiplied by the decrease in safety margin, where the first weight coefficient and the second weight coefficient are pre-set non-negative coefficients, and the decrease in safety margin is the absolute value when the change in safety margin is negative; if the change in safety margin is not negative, the decrease in safety margin is recorded as zero.

[0050] When extracting independent sets, first sort all nodes in the undirected graph of action conflicts from highest to lowest node weight. Then, starting with the nodes at the top of the sort, sequentially select nodes that have no edges between them and the selected nodes, adding them to the current independent set until no more nodes can be added, resulting in an independent set. After removing or marking the nodes in the current independent set and their selection states related to the sorting, repeat the above process to obtain multiple independent sets. For each independent set, sum the weights of all nodes it contains to obtain the total node weight corresponding to that independent set. Then, sort the independent sets from highest to lowest total node weight, and select the top-ranked independent sets as the candidate set of adjustment action combinations.

[0051] S7. Based on the improvement amount of each candidate adjustment action combination in the candidate adjustment action combination set to the main control target and the change amount to the boiler operation safety margin, determine the target adjustment action combination, and control the circulating fluidized bed boiler to perform adjustment according to the target adjustment action combination.

[0052] Specifically, each candidate control action combination in the candidate control action combination set is predicted one by one at the combination level to obtain the predicted bed temperature, predicted bed pressure difference, predicted furnace outlet oxygen, predicted main steam pressure, predicted coal feed rate, predicted primary air volume, predicted secondary air volume, and predicted return valve opening for each candidate control action combination. Based on this, the improvement amount of the main control target and the change amount of boiler operation safety margin corresponding to each candidate control action combination are calculated.

[0053] The improvement amount of the main control target is determined based on the type of main control target in the current control cycle. When the main control target is to restore the bed pressure differential, the improvement amount is determined based on the increase in the predicted bed pressure differential relative to the current bed pressure differential, or based on the reduction in the deviation of the bed pressure differential from the target lower limit. When the main control target is to reduce the bed temperature, the improvement amount is determined based on the decrease in the predicted bed temperature relative to the current bed temperature, or based on the reduction in the excess of the bed temperature relative to the preset upper limit boundary. When the main control target is to correct the main steam pressure, the improvement amount is determined based on the decrease in the absolute value of the predicted main steam pressure deviation relative to the absolute value of the current main steam pressure deviation. When the main control target is to correct the furnace outlet oxygen content, the improvement amount is determined based on the decrease in the deviation of the predicted furnace outlet oxygen content from the nearest boundary of the preset target interval, or based on the decrease in the absolute value of the deviation of the predicted furnace outlet oxygen content from the center value of the target interval.

[0054] In one embodiment, the change in boiler operating safety margin is the difference between the predicted boiler operating safety margin corresponding to the candidate adjustment action combination and the boiler operating safety margin in the current control cycle. When the change in boiler operating safety margin is negative, it indicates a decrease in boiler operating safety margin; when the change in boiler operating safety margin is positive, it indicates an increase in boiler operating safety margin. Candidate adjustment action combinations with a change in boiler operating safety margin less than a preset lower limit are first eliminated. Then, from the remaining candidate adjustment action combinations, the candidate adjustment action combination with the largest improvement in the main control target is selected as the target adjustment action combination. When two or more candidate adjustment action combinations have the same improvement in the main control target, the candidate adjustment action combination with the larger change in boiler operating safety margin is preferentially selected.

[0055] In another embodiment, the improvement amount of the main control target and the change in the boiler operating safety margin can be normalized separately and then weighted and summed to obtain a combined comprehensive score. The candidate adjustment action combination with the highest comprehensive score is determined as the target adjustment action combination, where the improvement amount of the main control target corresponds to a positive score item, and the decrease in the boiler operating safety margin corresponds to a negative score item. The preset lower limit and the weights of each score can be determined based on the boiler operating procedures, alarm thresholds, historical operating data, and adjustment effect calibration results.

[0056] Once the target adjustment action combination is determined, each discrete adjustment action is converted into an adjustment command with a corresponding execution amount. In one embodiment, the conversion method involves generating coal feeding adjustment commands, primary air adjustment commands, secondary air adjustment commands, and return valve opening adjustment commands according to the preset action step length corresponding to each discrete adjustment action, and sending them to the corresponding actuators to complete the adjustment. The preset action step length can be determined based on the actuator response speed, control cycle length, and the single-cycle adjustment amplitude allowed by the operating procedure.

[0057] Please see Figure 3 , Figure 3 This is a schematic diagram of the autonomous combustion optimization control system for a circulating fluidized bed boiler provided by the present invention. The present invention also provides an autonomous combustion optimization control system for a circulating fluidized bed boiler, used to implement the above method, including: The operating parameter acquisition module is used to collect the operating parameters of the circulating fluidized bed boiler during the current control cycle. The main control target determination module is used to determine the main control target for the current control cycle based on the operating parameters. The discrete action generation module is used to generate a set of corresponding discrete adjustment actions for the boiler combustion adjustment execution amount based on the operating parameters and the main control target. The constraint filtering module is used to substitute the operating parameters into a pre-established boiler combustion control constraint template, perform constraint filtering on the discrete adjustment action set, and obtain the set of allowed adjustment actions. The conflict graph construction module is used to determine the conflict between any two discrete adjustment actions in the set of allowed adjustment actions, and to construct an undirected graph of action conflicts based on the conflict determination results. The candidate combination extraction module is used to extract independent sets from the action conflict undirected graph to obtain a set of candidate adjustment action combinations. The target execution module is used to determine the target adjustment action combination based on the improvement amount of each candidate adjustment action combination in the candidate adjustment action combination set to the main control target and the change amount to the boiler operation safety margin, and to control the circulating fluidized bed boiler to perform adjustment according to the target adjustment action combination.

[0058] 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 a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0059] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for autonomous optimization control of combustion in a circulating fluidized bed boiler, characterized in that, include: S1. Collect the operating parameters of the circulating fluidized bed boiler during the current control cycle; S2. Determine the main control target for the current control cycle based on the operating parameters; S3. Based on the operating parameters and the main control target, generate a set of corresponding discrete adjustment actions for the boiler combustion adjustment execution amount; S4. Substitute the operating parameters into the pre-established boiler combustion control constraint template, perform constraint screening on the discrete adjustment action set, and obtain the allowable adjustment action set; S5. Perform conflict determination on any two discrete adjustment actions in the set of allowed adjustment actions, and construct an undirected graph of action conflicts based on the conflict determination results; S6. Extract independent sets from the undirected graph of action conflict to obtain a set of candidate adjustment action combinations; S7. Based on the improvement amount of each candidate adjustment action combination in the candidate adjustment action combination set to the main control target and the change amount to the boiler operation safety margin, determine the target adjustment action combination, and control the circulating fluidized bed boiler to perform adjustment according to the target adjustment action combination.

2. The autonomous optimization control method for combustion in a circulating fluidized bed boiler according to claim 1, characterized in that, The operating parameters include bed temperature, bed pressure difference, furnace outlet oxygen content, main steam pressure, coal feed rate, primary air volume, secondary air volume, and return valve opening.

3. The autonomous optimization control method for combustion in a circulating fluidized bed boiler according to claim 2, characterized in that, S2 includes: When the bed pressure difference is less than the first warning threshold, restoring the bed pressure difference will be determined as the main control target; When the bed pressure difference is not less than the first warning threshold and the bed temperature is greater than the second warning threshold, reducing the bed temperature will be determined as the main control target. When the bed pressure difference is not less than the first warning threshold, the bed temperature is not greater than the second warning threshold, and the main steam pressure deviation is greater than the third threshold, the corrected main steam pressure will be determined as the main control target. If none of the aforementioned conditions are met, the oxygen content at the furnace outlet will be determined as the primary control target.

4. The autonomous optimization control method for combustion in a circulating fluidized bed boiler according to claim 1, characterized in that, The boiler combustion regulation parameters include coal feed rate, primary air volume, secondary air volume, and return valve opening; S3 includes: The system generates actions to increase and decrease coal feed based on the coal feed rate, actions to increase and decrease primary air volume based on the primary air volume, actions to increase and decrease secondary air volume based on the secondary air volume, and actions to increase and decrease the return valve opening based on the return valve opening degree.

5. The autonomous optimization control method for combustion in a circulating fluidized bed boiler according to claim 1, characterized in that, The boiler combustion control constraint template includes bed temperature constraint, bed pressure difference constraint, furnace outlet oxygen constraint, main steam pressure constraint, coal feed variation constraint, air-coal matching constraint, staged air distribution constraint, and return material coordination constraint.

6. The autonomous optimization control method for combustion in a circulating fluidized bed boiler according to claim 1, characterized in that: The conflict determination results include the determination results corresponding to physical mutual exclusion relationships; The physical mutual exclusion relationship includes the relationship that the increase action and the decrease action corresponding to the same boiler combustion regulation execution amount cannot be realized at the same time within the same control cycle.

7. The autonomous optimization control method for combustion in a circulating fluidized bed boiler according to claim 1, characterized in that: The conflict determination results include the determination results corresponding to the running boundary approximation relationship; The boiler operation safety margin is determined based on the first boundary distance between the bed temperature and the preset upper limit boundary, the second boundary distance between the bed pressure difference and the preset lower limit boundary, the third boundary distance between the furnace outlet oxygen quantity and the preset target interval boundary, and the fourth boundary distance between the main steam pressure and the preset lower limit boundary in the one-step prediction results corresponding to any two discrete adjustment actions. The operational boundary approximation relationship is determined based on the first boundary distance and the second boundary distance; When the first boundary distance is less than the first threshold and the second boundary distance is less than the second threshold, it is determined that there is a running boundary approximation relationship between any two discrete adjustment actions.

8. The autonomous optimization control method for combustion in a circulating fluidized bed boiler according to claim 1, characterized in that: The conflict determination results include the determination results corresponding to the imbalance relationship of combustion organization; The combustion organization imbalance includes the imbalance between air and coal, the imbalance between primary and secondary air distribution, and the imbalance between return material and coal feeding. The air-coal matching imbalance index corresponding to the air-coal matching imbalance relationship, the graded air distribution imbalance index corresponding to the primary air and secondary air graded air distribution imbalance relationship, and the return material coordination imbalance index corresponding to the return material and coal feeding coordination imbalance relationship are determined based on the one-step prediction results corresponding to any two discrete adjustment actions. The preset allowable range is determined by taking the intersection of the allowable range given in the boiler operating procedure, the historical steady-state operating condition statistical range, and the boundary range of the mechanism model; When any of the following imbalance indicators exceeds the preset allowable range: the air-coal matching imbalance indicator, the graded air distribution imbalance indicator, and the return material coordination imbalance indicator, it is determined that there is a combustion organization imbalance relationship between any two discrete adjustment actions.

9. The autonomous optimization control method for combustion in a circulating fluidized bed boiler according to claim 1, characterized in that, S6 includes: Assign node weights to each node in the undirected graph of action conflict, extract multiple independent sets based on the node weights, sort them according to the total node weights corresponding to each independent set, and determine the candidate set of adjustment action combinations. The node weights are determined based on the improvement of the main control objective by the corresponding discrete adjustment action and the change in the boiler operation safety margin.

10. An autonomous combustion optimization control system for a circulating fluidized bed boiler, characterized in that, include: The operating parameter acquisition module is used to collect the operating parameters of the circulating fluidized bed boiler during the current control cycle. The main control target determination module is used to determine the main control target for the current control cycle based on the operating parameters. The discrete action generation module is used to generate a set of corresponding discrete adjustment actions for the boiler combustion adjustment execution amount based on the operating parameters and the main control target. The constraint filtering module is used to substitute the operating parameters into a pre-established boiler combustion control constraint template, perform constraint filtering on the discrete adjustment action set, and obtain the set of allowed adjustment actions. The conflict graph construction module is used to determine the conflict between any two discrete adjustment actions in the set of allowed adjustment actions, and to construct an undirected graph of action conflicts based on the conflict determination results. The candidate combination extraction module is used to extract independent sets from the action conflict undirected graph to obtain a set of candidate adjustment action combinations. The target execution module is used to determine the target adjustment action combination based on the improvement amount of each candidate adjustment action combination in the candidate adjustment action combination set to the main control target and the change amount to the boiler operation safety margin, and to control the circulating fluidized bed boiler to perform adjustment according to the target adjustment action combination.

Images