Multi-objective optimization method and device for biomass-coal blending combustion

Abstract

The invention provides a multi-objective optimization method and device for biomass-coal blending combustion, and relates to the technical field of combustion optimizing.The method comprises the steps that biomass information and coal information are obtained and normalized and quantified into an information matrix; defining multiple evaluation dimensions, decomposing the information matrix, performing parallel construction and multi-target fitting, and determining a target optimization module; and aiming at the whole blending combustion period, dimension evaluation and integration based on a target optimization module are executed, blending ratio data are determined, and the whole blending combustion period is defined as multiple stages based on blending combustion intensity from the initial blending combustion stage to the complete burnout stage. The technical problems that in the prior art, due to the fact that collaborative optimization of biomass blending combustion and coal blending combustion is insufficient, multiple targets cannot be considered at the same time, and the operation requirements of a coal power unit under different working conditions and multiple fuel combinations are difficult to meet are solved, multi-target optimization is carried out on the whole period of blending combustion, and the operation efficiency of the coal power unit is improved. And the operation requirements of the coal power unit under different working conditions and various fuel combinations are met.

Description

Technical Field

[0001] This application relates to the field of combustion optimization technology, and in particular to a multi-objective optimization method and apparatus for biomass-coal co-firing. Background Technology

[0002] Biomass co-firing, as an effective energy-saving and carbon-reducing technology, often employs a blending strategy of high-quality and low-quality coal in coal-fired power units to ensure combustion stability due to differences in coal quality entering the furnace. Biomass-coal co-firing involves multiple objectives, including boiler operational safety, system low-carbon performance, and production ratio. However, existing research on biomass co-firing often focuses on boiler operational safety, neglecting the synergistic optimization of low-carbon performance and production ratio. While existing coal blending methods address the safety and production ratio of blended coal combustion, they fail to fully consider low-carbon objectives within the context of low-carbonization and lack comprehensive consideration of biomass co-firing, making it difficult to meet the operational needs of coal-fired power units under multiple operating conditions and fuel combinations.

[0003] In summary, existing technologies suffer from insufficient synergistic optimization between biomass co-firing and coal co-firing, resulting in the inability to simultaneously address multiple objectives and meet the operational requirements of coal-fired power units under different operating conditions and multiple fuel combinations. Summary of the Invention

[0004] The purpose of this application is to provide a multi-objective optimization method and apparatus for biomass-coal blending, in order to solve the technical problem in the prior art that the insufficient synergistic optimization of biomass blending and coal blending leads to the inability to simultaneously consider multiple objectives and meet the operational requirements of coal-fired power units under different operating conditions and multiple fuel combinations.

[0005] In view of the above problems, this application provides a multi-objective optimization method and apparatus for biomass-coal co-firing.

[0006] Firstly, this application provides a multi-objective optimization method for biomass-coal blending, which is implemented through a multi-objective optimization device for biomass-coal blending. The method includes: acquiring biomass and coal information, normalizing and quantizing them into an information matrix, wherein the information matrix includes a first information matrix of biomass information and a second information matrix of coal information; defining multiple evaluation dimensions and decomposing the information matrix, performing parallel construction and multi-objective fitting to determine the target optimization module, wherein the parallel construction includes a first evaluation model based on the low-carbon properties of biomass-coal blending, a second evaluation model based on the safety of biomass-coal blending, and a third evaluation model based on the biomass-coal blending input-output ratio; and performing dimensional evaluation and synthesis based on the target optimization module for the entire blending cycle to determine the blending ratio data, wherein the entire blending cycle from the initial stage of blending to complete combustion is defined as a multi-stage based on the blending intensity.

[0007] Optionally, information on pre-blended biomass and coal is obtained, and a first element matrix and a second element matrix are determined by extracting combustion-related elements; the first element matrix and the second element matrix are then normalized and quantized to integrate and determine the information matrix.

[0008] Optionally, the first element matrix and the second element matrix are traversed to determine the normalized dimensions, wherein the normalized dimensions correspond to the combustion-related elements; the first element matrix and the second element matrix are matched for common elements and processed with the normalized dimensions to determine the first information matrix and the second information matrix; the first information matrix and the second information matrix are integrated to determine the information matrix.

[0009] Optionally, the information matrix is ​​filtered based on the low carbon content of biomass-coal co-firing to determine a one-dimensional matrix; the information matrix is ​​filtered based on the safety of biomass-coal co-firing to determine a two-dimensional matrix; the information matrix is ​​filtered based on the biomass-coal co-firing production ratio to determine a three-dimensional matrix; and the target optimization module is constructed based on the one-dimensional matrix, the two-dimensional matrix, and the three-dimensional matrix.

[0010] Optionally, a first evaluation rule is introduced; by embedding the one-dimensional matrix, the first evaluation model is constructed based on the first evaluation rule and using a sample-driven training method.

[0011] Optionally, the first evaluation model, the second evaluation model, and the third evaluation model are implemented in parallel and then connected to a multi-objective solution model to form the objective optimization module.

[0012] Optionally, the multi-objective solution model is constructed; wherein the outputs of the first evaluation model, the second evaluation model, and the third evaluation model are used as input data, multi-objective equilibrium is used as a constraint, and the blending ratio is used as output data to supervise the training of the multi-objective solution model.

[0013] Optionally, the entire co-firing cycle is divided into stages based on the co-firing intensity to determine the co-firing stage sequence; for the co-firing stage sequence, a single-dimensional evaluation and a comprehensive evaluation are triggered based on the target optimization module to determine the co-firing ratio data.

[0014] Optionally, a mixer is connected, and the blending ratio data is transmitted to the central control system of the mixer according to a communication protocol; the central control system interprets the blending ratio data and drives the first feeding port, the second feeding port and the mixing components to perform blending control, wherein the first feeding port performs biomass feeding.

[0015] Secondly, this application also provides a multi-objective optimization device for biomass-coal blending, used to execute a multi-objective optimization method for biomass-coal blending as described in the first aspect. The multi-objective optimization device for biomass-coal blending includes: an information matrix normalization module, used to acquire biomass information and coal information, normalize and quantize them into an information matrix, wherein the information matrix includes a first information matrix of biomass information and a second information matrix of coal information; a parallel construction module, used to define multiple evaluation dimensions and decompose the information matrix, perform parallel construction and multi-objective fitting, and determine the target optimization module, wherein the parallel construction includes a first evaluation model based on the low-carbon properties of biomass-coal blending, a second evaluation model based on the safety of biomass-coal blending, and a third evaluation model based on the biomass-coal blending production ratio; and a full-cycle evaluation module, used to perform dimensional evaluation and synthesis based on the target optimization module for the entire blending cycle, and determine the blending ratio data, wherein the entire blending cycle from the initial stage of blending to complete combustion is defined as a multi-stage based on the blending intensity.

[0016] One or more technical solutions provided in this application have at least the following beneficial effects:

[0017] By acquiring biomass and coal information, normalizing and quantifying it into an information matrix, wherein the information matrix includes a first information matrix of biomass information and a second information matrix of coal information; defining multiple evaluation dimensions and decomposing the information matrix, performing parallel construction and multi-objective fitting, and determining the target optimization module, wherein the parallel construction includes a first evaluation model based on the low carbon nature of biomass-coal blending, a second evaluation model based on the safety of biomass-coal blending, and a third evaluation model based on the biomass-coal blending production ratio; for the entire blending cycle, performing dimensional evaluation and synthesis based on the target optimization module to determine the blending ratio data, wherein the entire blending cycle from the initial stage of blending to complete combustion is defined as a multi-stage based on the blending intensity. In other words, by normalizing and matrixing the complex information of biomass and coal, defining multiple evaluation dimensions, and performing multi-objective optimization for the entire co-firing cycle, an executable biomass / coal co-firing ratio is output under the premise of meeting boiler safety constraints. This meets the operational needs of coal-fired power units under different operating conditions and multiple fuel combinations, improves co-firing efficiency, and effectively reduces carbon emissions per unit energy output while ensuring thermal efficiency and economy.

[0018] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating a multi-objective optimization method for biomass-coal co-firing according to this application.

[0021] Figure 2 This is a schematic diagram of the structure of a multi-objective optimization device for biomass-coal co-firing according to this application.

[0022] Figure labeling: Information matrix normalization module 11, parallel construction module 12, full-cycle evaluation module 13. Detailed Implementation

[0023] This application provides a multi-objective optimization method and apparatus for biomass-coal blending, solving the technical problem in existing technologies where insufficient synergistic optimization of biomass and coal blending leads to the inability to simultaneously consider multiple objectives, making it difficult to meet the operational needs of coal-fired power units under different operating conditions and multiple fuel combinations. By normalizing and matrixing the complex information of biomass and coal, defining multiple evaluation dimensions, and performing multi-objective optimization throughout the blending cycle, this method outputs an executable biomass / coal blending ratio while meeting boiler safety constraints. This satisfies the operational needs of coal-fired power units under different operating conditions and multiple fuel combinations, improves blending efficiency, and effectively reduces carbon emissions per unit energy output while ensuring thermal efficiency and economy.

[0024] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0025] Example 1, please refer to the appendix. Figure 1 This application provides a multi-objective optimization method for biomass-coal blending, wherein the multi-objective optimization method for biomass-coal blending is executed by a multi-objective optimization device for biomass-coal blending, and the multi-objective optimization method for biomass-coal blending specifically includes the following steps:

[0026] S100: Obtain biomass information and coal information, normalize and quantize them into an information matrix, wherein the information matrix includes a first information matrix of biomass information and a second information matrix of coal information.

[0027] Furthermore, this application S100 includes:

[0028] Information on pre-blended biomass and coal is obtained, and combustion-related elements are extracted to determine a first element matrix and a second element matrix. The first element matrix and the second element matrix are then normalized and quantized to integrate and determine the information matrix.

[0029] Furthermore, this application also includes the following steps:

[0030] Traverse the first element matrix and the second element matrix to determine the normalized dimensions, wherein the normalized dimensions correspond to combustion-related elements; perform source element matching on the first element matrix and the second element matrix, and process them with the normalized dimensions to determine the first information matrix and the second information matrix; integrate the first information matrix and the second information matrix to determine the information matrix.

[0031] Specifically, basic information on biomass and coal is collected, including calorific value Q (in MJ / kg); carbon content C (in %), representing the carbon content of a unit mass of biomass or coal; volatile matter V (in %), representing the volatile matter content of a unit mass of biomass or coal; sulfur content S (in %), representing the sulfur content of a unit mass of biomass or coal; moisture content M (in %), representing the moisture content of a unit mass of biomass or coal; ash content A (in %), representing the ash content of a unit mass of biomass or coal; and unit price P (in yuan / ton), representing the price of a unit mass of biomass or coal.

[0032] Biomass information refers to the physicochemical properties of biomass fuels, such as calorific value, carbon content, volatile matter, sulfur content, moisture content, ash content, and price. Coal information refers to similar property data for coal, including calorific value, carbon content, volatile matter, sulfur content, moisture content, ash content, and price. Parameters closely related to combustion performance (such as calorific value, volatile matter, and carbon content) are extracted from the biomass and coal information to obtain a first element matrix and a second element matrix. The first element matrix is ​​constructed based on the elements extracted from the biomass information, and the second element matrix is ​​constructed based on the elements extracted from the coal information. Each row of each matrix represents a fuel sample, and each column represents a specific element related to combustion performance.

[0033] For each element in the first and second element matrices, normalization is performed to ensure that different types of elements (such as carbon content, calorific value, ash content, etc.) are compared under the same standard. For example, the min-max normalization method is used to convert all data into a range of 0 to 1. The same elements in the first and second element matrices are then matched. For example, the carbon content C of biomass is matched with the carbon content C of coal, and the volatile matter V is matched with the volatile matter V. These elements are then uniformly processed by normalizing their dimensions to obtain the first and second information matrices.

[0034] By integrating the first and second information matrices, an information matrix is ​​obtained, containing normalized data for various combustion performance elements. Through normalization and matrix processing, complex biomass and coal information is unified into standardized data, avoiding errors caused by inconsistencies in data units and dimensions, and improving the accuracy and comparability of data processing.

[0035] S200: Define multiple evaluation dimensions and decompose the information matrix, perform parallel construction and multi-objective fitting, and determine the objective optimization module. The parallel construction includes a first evaluation model based on the low carbon nature of biomass-coal co-firing, a second evaluation model based on the safety of biomass-coal co-firing, and a third evaluation model based on the biomass-coal co-firing production ratio.

[0036] Furthermore, this application S200 includes:

[0037] The information matrix is ​​filtered based on the low-carbon properties of biomass-coal co-firing to determine a one-dimensional matrix; the information matrix is ​​filtered based on the safety of biomass-coal co-firing to determine a two-dimensional matrix; the information matrix is ​​filtered based on the biomass-coal co-firing production ratio to determine a three-dimensional matrix; and the target optimization module is constructed based on the one-dimensional, two-dimensional, and three-dimensional matrices.

[0038] Furthermore, this application also includes the following steps:

[0039] A first evaluation rule is introduced; the first evaluation model is constructed by embedding the one-dimensional matrix and using the first evaluation rule as a benchmark, through sample-driven training.

[0040] The first, second, and third evaluation models are implemented in parallel, and then connected to a multi-objective solution model to form the objective optimization module.

[0041] Construct the multi-objective solution model; wherein the outputs of the first evaluation model, the second evaluation model and the third evaluation model are used as input data, multi-objective equilibrium is used as constraint, and the blending ratio is used as output data, and the multi-objective solution model is trained under supervision.

[0042] Specifically, the low-carbon nature of biomass-coal co-combustion refers to the characteristics of reducing greenhouse gas (mainly carbon dioxide) and other pollutants (such as nitrogen oxides NOx and sulfur oxides SOx) emissions during the co-combustion of biomass and coal. It focuses on the degree of environmental impact of the combustion process and is one of the optimization goals. The safety of biomass-coal co-combustion refers to the characteristics of ensuring the stable and reliable operation of boilers and related equipment during the co-combustion of biomass and coal, avoiding dangerous situations (such as fire extinguishing, deflagration, slagging, corrosion, and high-temperature corrosion). It focuses on risk control during the combustion process and is another optimization goal. The input-output ratio of biomass-coal co-combustion can be understood as the input-output ratio. It focuses on the proportional relationship between the total cost of the co-combustion process (including the purchase cost of biomass and coal, processing cost, operation and maintenance cost, etc.) and its benefits (including power generation revenue, reduced carbon emission costs, and potential subsidies, etc.). Pursuing a higher input-output ratio means more economical and efficient operation, which is the third optimization goal.

[0043] The information matrix is ​​filtered based on the low-carbon performance of biomass-coal blending, the safety of biomass-coal blending, and the input-output ratio of biomass-coal blending, extracting the data subsets most relevant to each objective, namely one-dimensional, two-dimensional, and three-dimensional matrices. The originally complex multi-objective problem is decomposed into data processing steps related to each individual objective (or a combination of main objectives). The one-dimensional matrix is ​​mainly input into the low-carbon performance assessment model, the two-dimensional matrix into the safety assessment model, and the three-dimensional matrix into the economic performance assessment model.

[0044] The first assessment rule is introduced, which is used to assess the low-carbon performance of fuel blends. The first assessment model is a biomass-coal blending low-carbon performance assessment model, used to evaluate the low-carbon performance of blended biomass and coal, assuming a total biomass content of n... b Types, coal total n c species, x i Let L represent the proportion of each type of coal or biomass. Then, the low-carbon assessment index L for biomass-coal co-firing is: Let represent the objective function, where This indicates the carbon quota allocated to the generating unit. This represents the carbon dioxide emissions from the combustion of biomass-coal fuel. Denotes the constraint condition, where FC i C represents the consumption of the i-th type of fuel, in tons (t); i OF represents the carbon content of the base element of the i-th fuel, expressed in %; i This represents the carbon oxidation rate of the i-th fuel, expressed as a percentage.

[0045] The previously generated one-dimensional matrix (i.e., a matrix containing carbon emission intensity data) is input into the model as training data. Each matrix element represents the carbon emission intensity of a specific fuel combination, serving as the basis for evaluation. Using the one-dimensional matrix, the model is trained according to the first evaluation rule, enabling it to predict low-carbon performance based on input features (such as fuel composition, calorific value, carbon content, etc.). During training, the model adjusts its internal parameters to predict the target assessment (such as low carbon emissions) as accurately as possible given the input features. After training, a first evaluation model is constructed that can predict low-carbon performance based on parameters such as the blending ratio of biomass and coal and carbon content.

[0046] Similarly, the second evaluation model is a biomass-coal co-combustion safety assessment model, used to evaluate the safety of biomass-coal co-combustion. Volatile matter is the driving force of combustion; high volatile matter promotes ignition and combustion rate. Ash content is a resistance to combustion, increasing the risk of slagging and reducing thermal efficiency. Moisture evaporation absorbs heat, lowering combustion temperature; high moisture content increases pulverization difficulty and can clog pipelines. Based on the above parameter characteristics of biomass-coal mixtures, a comprehensive combustion performance index (CPI) is proposed: Let V represent the objective function, where V, A, and M represent the volatile matter, ash content, and moisture content of the biomass-coal blend, respectively. The calculation formula is as follows: Denotes the constraint condition, where V i A i M i Let represent the volatile matter, ash content, and moisture content of the i-th fuel, respectively; k represents the moisture influence coefficient, which is the ratio of the inhibitory effect of moisture on combustion to the inhibitory effect of ash. To ensure that the combustion performance of the boiler after biomass-coal co-firing meets the standards, the following constraints need to be kept within a reasonable range:

[0047]

[0048] This represents the constraint conditions, where the maximum and minimum values ​​of each parameter are the boiler design requirements.

[0049] The previously generated two-dimensional matrix is ​​used as training data and input into the model. Using this matrix, the model is trained according to a second evaluation rule, enabling it to predict safety based on input features (such as fuel composition, calorific value, and carbon content). During training, the model adjusts its internal parameters to predict the target assessment as accurately as possible given the input features. After training, a second evaluation model is constructed that can predict safety performance based on parameters such as the blending ratio of biomass and coal and carbon content.

[0050] The third evaluation model is the biomass-coal blending production ratio evaluation model, used to evaluate the cost Z after blending biomass and coal: Let n represent the objective function, where n is the target function. b n represents the number of biomass species. c x represents the number of types of coal. i P represents the proportion of each type of coal or biomass. i The unit price of the i-th fuel. The previously generated 3D matrix is ​​input into the model as training data. Using the 3D matrix, the model is trained using a third evaluation rule, enabling it to predict the input-output ratio based on input features (such as fuel composition, calorific value, carbon content, etc.). During training, the model adjusts its internal parameters to predict the target evaluation as accurately as possible given the input features. After training, a second evaluation model is constructed that can predict the input-output ratio based on parameters such as the blending ratio of biomass and coal, and carbon content.

[0051] The first, second, and third evaluation models are connected in parallel, and their outputs are used as inputs to a unified multi-objective solution model. This model performs trade-off optimization on multiple objective functions, forming the objective optimization module. The objective optimization module is a complete structural module consisting of an evaluation submodule and a solution submodule, used to implement a closed-loop optimization system from raw data input and performance evaluation to optimal strategy output.

[0052] A multi-objective optimization model is an optimization model established for multiple conflicting or mutually restrictive objectives (such as low carbon emissions, safety, and input-output ratio). It uses algorithms to jointly solve for each objective function, outputting the blending ratio (the proportion of each fuel). A trade-off mechanism is set between the objective functions as constraints for optimization. For example, it may require that L not exceed a certain value, CPI not fall below a certain threshold, and Z be within the budget; or it may set a weight balance function w1·L + w2·(1 / CPI) + w3·Z. The output of the multi-objective optimization model is the blending ratio of each fuel (biomass and different coal types), which usually needs to meet the blending ratio and constraints (such as a total sum of 1, a certain fuel proportion not exceeding a limit, etc.).

[0053] Using the outputs of the first, second, and third evaluation models as input data, multi-objective equilibrium as constraint, and blending ratio as output data, a supervised learning model is trained to fit the functional relationship between the evaluation index and the blending ratio. During training, the loss function (e.g., MSE) is continuously optimized to verify the rationality of the model's output on unknown samples and to confirm whether the set multi-objective equilibrium constraints are met. The biomass-coal blending multi-objective solution model is used to summarize and solve the objective functions and constraints of biomass-coal blending, such as low carbon emissions, safety, and input-output ratio. In addition, there is also a constraint on the biomass-coal ratio. The constraints are represented, and the model is solved using the NSGA-II algorithm to obtain a Pareto optimal solution set. These solutions characterize the trade-offs between low carbon emissions, safety, and input-output ratio, and can determine the specific ratio of biomass and coal blending. This is used to summarize the three models and calculate different blending ratios under different Pareto boundaries using the built-in NSGA-II algorithm, providing this information to production personnel as a reference for actual blending.

[0054] By defining multiple evaluation dimensions and decomposing the information matrix, development efficiency was improved, interference between different objectives during model building was avoided, and the accuracy of each model evaluation was guaranteed. The determination of multi-objective fitting and objective optimization modules enabled a leap from multiple independent evaluations to comprehensive optimization. This allows for the balancing of sometimes conflicting objectives within a unified framework, ensuring that the final optimization result is no longer a partial optimum of a single objective, thus greatly enhancing the effectiveness and practicality of the optimization method.

[0055] S300: For the entire co-firing cycle, perform dimensional evaluation and integration based on the target optimization module to determine the co-firing ratio data. The entire co-firing cycle from the initial stage of co-firing to complete combustion is defined as a multi-stage based on the co-firing intensity.

[0056] Furthermore, this application S300 includes:

[0057] Based on the co-firing intensity, the entire co-firing cycle is divided into stages to determine the co-firing stage sequence; for the co-firing stage sequence, a single-dimensional evaluation and a comprehensive evaluation based on the target optimization module are triggered to determine the blending ratio data.

[0058] Furthermore, this application also includes the following steps:

[0059] The mixer is connected, and the blending ratio data is transmitted to the central control system of the mixer according to the communication protocol. The central control system interprets the blending ratio data and drives the first feeding port, the second feeding port and the mixing components to perform blending control. The first feeding port performs biomass feeding.

[0060] Specifically, blending intensity refers to the proportions of biomass and coal used in the blending process and their impact on combustion efficiency and emissions. Blending intensity can be quantified by multiple factors (such as the calorific value and carbon content of biomass) to reflect the contribution of the biomass-coal blending combination to the overall combustion process. The entire blending cycle is the entire process from the initial biomass-coal fuel ratio to the complete combustion, including the initial combustion stage (biomass ignition stage), the stable combustion stage (combustion sustaining stage), and the final combustion stage (complete burnout stage).

[0061] Based on the changes in co-firing intensity, the co-firing process is divided into multiple stages. The co-firing stage sequence includes all stages from the initial stage of co-firing to complete combustion. Each stage is defined and evaluated according to its combustion characteristics (such as temperature, pressure, fuel consumption, etc.). At the beginning of each stage, the target optimization module is triggered to perform single-dimensional and comprehensive evaluations. Single-dimensional evaluation assesses the co-firing combination from a single dimension (such as carbon emissions, safety, or cost); comprehensive evaluation, within a multi-objective framework, comprehensively considers multiple evaluation results such as low carbon emissions, safety, and input-output ratio, ultimately making a comprehensive optimization decision. A specific target evaluation model is executed independently for each stage. For example, in the initial stage of co-firing, the focus is on the low carbon emissions evaluation model to ensure that biomass participates in combustion as much as possible; in the stable combustion stage, more attention is paid to the safety evaluation model to ensure that the combustion process does not cause safety hazards. Based on the target optimization module, a comprehensive evaluation is performed considering multiple objectives (low carbon emissions, safety, input-output ratio), outputting optimized co-firing ratio data for all stages.

[0062] Based on the phased evaluation results, and combined with a multi-objective optimization algorithm (such as NSGA-II), the optimal blending ratio is determined. The optimization results must meet the following requirements: under the low-carbon objective, maintain a relatively high biomass blending ratio; under the safety objective, ensure fuel stability and efficiency during combustion; and under the production ratio objective, minimize the unit calorific value cost. The blending ratio data is the optimal biomass-coal blending ratio calculated based on the evaluation results. It is determined based on the results of each phase of the evaluation, with the goal of achieving optimal low-carbon performance, safety, and production ratio throughout the entire combustion cycle.

[0063] A connection is established with the mixer, and the blending ratio data (e.g., 30% biomass, 70% coal) obtained from the target optimization module is transmitted to the mixer's central control system via a communication protocol (such as industrial protocols like Modbus and OPC). The communication protocol ensures accurate data transmission from the calculation module to the mixer, guaranteeing error-free and packet-free data transmission. For example, data can be sent to the central control system via TCP / IP, in JSON or XML format, including the biomass to coal ratio: e.g., 0.3 biomass, 0.7 coal. Upon receiving the blending ratio data, the central control system interprets this data and converts it into specific operational instructions: based on the data (e.g., 30% biomass), it controls the first feed port (biomass feed port) to ensure sufficient biomass supply. Based on the data (e.g., 70% coal), it controls the second feed port (coal feed port) to ensure sufficient coal supply. The operating mode of the mixing components is adjusted according to the blending ratio to ensure thorough mixing of the materials. The feed port is the material inlet in the mixer, responsible for separately feeding biomass and coal into the mixing chamber. Each feed port corresponds to a material type, such as a biomass feed port or a coal feed port. The mixing component is the drive and mixing device inside the mixer, used to mix different materials in a specified ratio.

[0064] The central control system sends a control signal to open the first feeding port, allowing biomass to be added to the mixer at a set ratio (e.g., 30%). The feeding rate and quantity are precisely adjusted according to the blending ratio calculated by the target optimization module, ensuring an accurate ratio of biomass to coal. Through multi-stage division based on co-firing intensity, each stage has its specific combustion characteristics and emission performance, requiring different optimization strategies. By combining single-dimensional and comprehensive evaluations, the optimal blending ratio data can be determined while meeting the requirements of multiple evaluation dimensions, improving the stability of the co-firing process. Automated control ensures that biomass and coal are accurately and uniformly mixed according to the optimized blending ratio data, avoiding errors and fluctuations that may occur during manual operation. This not only improves mixing efficiency and quality but also helps stabilize the boiler combustion process, reduces pollutant emissions, lowers labor costs, and increases production efficiency.

[0065] In summary, the multi-objective optimization method for biomass-coal co-firing provided in this application has the following beneficial effects:

[0066] By acquiring biomass and coal information, normalizing and quantifying it into an information matrix, wherein the information matrix includes a first information matrix of biomass information and a second information matrix of coal information; defining multiple evaluation dimensions and decomposing the information matrix, performing parallel construction and multi-objective fitting, and determining the target optimization module, wherein the parallel construction includes a first evaluation model based on the low carbon nature of biomass-coal blending, a second evaluation model based on the safety of biomass-coal blending, and a third evaluation model based on the biomass-coal blending production ratio; for the entire blending cycle, performing dimensional evaluation and synthesis based on the target optimization module to determine the blending ratio data, wherein the entire blending cycle from the initial stage of blending to complete combustion is defined as a multi-stage based on the blending intensity. In other words, by normalizing and matrixing the complex information of biomass and coal, defining multiple evaluation dimensions, and performing multi-objective optimization for the entire co-firing cycle, an executable biomass / coal co-firing ratio is output under the premise of meeting boiler safety constraints. This meets the operational needs of coal-fired power units under different operating conditions and multiple fuel combinations, improves co-firing efficiency, and effectively reduces carbon emissions per unit energy output while ensuring thermal efficiency and economy.

[0067] Example 2: Based on the same inventive concept as the multi-objective optimization method for biomass-coal blending in Example 1, this application also provides a multi-objective optimization device for biomass-coal blending. Please refer to the appendix. Figure 2 The aforementioned multi-objective optimization device for biomass-coal co-firing includes:

[0068] The information matrix normalization module 11 is used to acquire biomass information and coal information, normalize and quantify them into an information matrix, wherein the information matrix includes a first information matrix of biomass information and a second information matrix of coal information; the parallel construction module 12 is used to define multiple evaluation dimensions and decompose the information matrix, perform parallel construction and multi-objective fitting, and determine the target optimization module, wherein the parallel construction includes a first evaluation model based on the low carbon nature of biomass-coal blending, a second evaluation model based on the safety of biomass-coal blending, and a third evaluation model based on the biomass-coal blending production ratio; the full-cycle evaluation module 13 is used to perform dimensional evaluation and synthesis based on the target optimization module for the entire blending cycle, and determine the blending ratio data, wherein the entire blending cycle from the initial stage of blending to complete combustion is defined as a multi-stage based on the blending intensity.

[0069] Furthermore, the information matrix normalization module 11 in the aforementioned biomass-coal co-firing multi-objective optimization device is also used for:

[0070] Information on pre-blended biomass and coal is obtained, and combustion-related elements are extracted to determine a first element matrix and a second element matrix. The first element matrix and the second element matrix are then normalized and quantized to integrate and determine the information matrix.

[0071] Furthermore, the information matrix normalization module 11 in the aforementioned biomass-coal co-firing multi-objective optimization device is also used for:

[0072] Traverse the first element matrix and the second element matrix to determine the normalized dimensions, wherein the normalized dimensions correspond to combustion-related elements; perform source element matching on the first element matrix and the second element matrix, and process them with the normalized dimensions to determine the first information matrix and the second information matrix; integrate the first information matrix and the second information matrix to determine the information matrix.

[0073] Furthermore, the parallel construction module 12 in the aforementioned biomass-coal co-firing multi-objective optimization device is also used for:

[0074] The information matrix is ​​filtered based on the low-carbon properties of biomass-coal co-firing to determine a one-dimensional matrix; the information matrix is ​​filtered based on the safety of biomass-coal co-firing to determine a two-dimensional matrix; the information matrix is ​​filtered based on the biomass-coal co-firing production ratio to determine a three-dimensional matrix; and the target optimization module is constructed based on the one-dimensional, two-dimensional, and three-dimensional matrices.

[0075] Furthermore, the parallel construction module 12 in the aforementioned biomass-coal co-firing multi-objective optimization device is also used for:

[0076] A first evaluation rule is introduced; the first evaluation model is constructed by embedding the one-dimensional matrix and using the first evaluation rule as a benchmark, through sample-driven training.

[0077] Furthermore, the parallel construction module 12 in the aforementioned biomass-coal co-firing multi-objective optimization device is also used for:

[0078] The first, second, and third evaluation models are implemented in parallel, and then connected to a multi-objective solution model to form the objective optimization module.

[0079] Furthermore, the parallel construction module 12 in the aforementioned biomass-coal co-firing multi-objective optimization device is also used for:

[0080] Construct the multi-objective solution model; wherein the outputs of the first evaluation model, the second evaluation model and the third evaluation model are used as input data, multi-objective equilibrium is used as constraint, and the blending ratio is used as output data, and the multi-objective solution model is trained under supervision.

[0081] Furthermore, the full-cycle evaluation module 13 in the aforementioned biomass-coal co-firing multi-objective optimization device is also used for:

[0082] Based on the co-firing intensity, the entire co-firing cycle is divided into stages to determine the co-firing stage sequence; for the co-firing stage sequence, a single-dimensional evaluation and a comprehensive evaluation based on the target optimization module are triggered to determine the blending ratio data.

[0083] Furthermore, the full-cycle evaluation module 13 in the aforementioned biomass-coal co-firing multi-objective optimization device is also used for:

[0084] The mixer is connected, and the blending ratio data is transmitted to the central control system of the mixer according to the communication protocol. The central control system interprets the blending ratio data and drives the first feeding port, the second feeding port and the mixing components to perform blending control. The first feeding port performs biomass feeding.

[0085] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Figure 1 The multi-objective optimization method and specific example of biomass-coal co-firing in Example 1 are also applicable to the multi-objective optimization device of biomass-coal co-firing in this example. Through the foregoing detailed description of the multi-objective optimization method of biomass-coal co-firing, those skilled in the art can clearly understand the multi-objective optimization device of biomass-coal co-firing in this example. Therefore, for the sake of brevity, it will not be described in detail here.

[0086] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0087] Obviously, those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A multi-objective optimization method for biomass-coal co-firing, characterized in that, include: Biomass information and coal information are acquired, normalized and quantized into an information matrix, wherein the information matrix includes a first information matrix of biomass information and a second information matrix of coal information; Define multiple evaluation dimensions and decompose the information matrix, perform parallel construction and multi-objective fitting, and determine the target optimization module. The parallel construction includes a first evaluation model based on the low carbon nature of biomass-coal co-firing, a second evaluation model based on the safety of biomass-coal co-firing, and a third evaluation model based on the input-output ratio of biomass-coal co-firing. For the entire co-firing cycle, a dimensional evaluation and synthesis based on the target optimization module is performed to determine the blending ratio data. The entire co-firing cycle from the initial stage of co-firing to complete combustion is defined as a multi-stage process based on the co-firing intensity.

2. The multi-objective optimization method for biomass-coal blending as described in claim 1, characterized in that, Information on biomass and coal was acquired, normalized, and quantified into an information matrix, including: Information on pre-blended biomass and coal was obtained, and the first and second element matrices were determined by extracting combustion-related elements. The first element matrix and the second element matrix are normalized and quantized to integrate and determine the information matrix.

3. The multi-objective optimization method for biomass-coal co-firing as described in claim 2, characterized in that, Normalization and quantization processing are performed to integrate and determine the information matrix, including: Traverse the first element matrix and the second element matrix to determine the normalized dimensions, where the normalized dimensions correspond to the combustion-related elements; Perform source element matching on the first element matrix and the second element matrix, and process them with the normalized dimensions to determine the first information matrix and the second information matrix; The information matrix is ​​determined by integrating the first information matrix and the second information matrix.

4. The multi-objective optimization method for biomass-coal blending as described in claim 1, characterized in that, The target optimization module is defined, including: The information matrix is ​​filtered based on the low-carbon properties of biomass-coal co-firing to determine a one-dimensional matrix. The information matrix is ​​filtered based on the safety of biomass-coal co-firing to determine a two-dimensional matrix. The information matrix is ​​filtered based on the biomass-coal blending ratio to determine the three-dimensional matrix; The target optimization module is constructed based on the one-dimensional matrix, the two-dimensional matrix, and the three-dimensional matrix.

5. The multi-objective optimization method for biomass-coal blending as described in claim 4, characterized in that, Introduce the first evaluation rule; By embedding the one-dimensional matrix and using the first evaluation rule as a benchmark, the first evaluation model is constructed using a sample-driven training method.

6. The multi-objective optimization method for biomass-coal co-firing as described in claim 5, characterized in that, The first, second, and third evaluation models are implemented in parallel, and then connected to a multi-objective solution model to form the objective optimization module.

7. The multi-objective optimization method for biomass-coal blending as described in claim 6, characterized in that, Construct the multi-objective solution model; The multi-objective solution model is trained under supervision by using the outputs of the first, second, and third evaluation models as input data, multi-objective equilibrium as constraint, and blending ratio as output data.

8. The multi-objective optimization method for biomass-coal blending as described in claim 1, characterized in that, Perform dimensional evaluation and synthesis based on the target optimization module to determine the blending ratio data, including: Based on the blending intensity, the entire blending cycle is divided into stages to determine the blending stage sequence; For the aforementioned co-firing stage sequence, a single-dimensional evaluation and a comprehensive evaluation are triggered based on the target optimization module to determine the co-firing ratio data.

9. The multi-objective optimization method for biomass-coal blending as described in claim 1, characterized in that, After determining the blending ratio data, the following is included: Connect the mixer and, according to the communication protocol, transmit the blending ratio data to the central control system of the mixer; The central control system interprets the blending ratio data to control and drive the blending of the first feeding port, the second feeding port, and the mixing component, wherein the first feeding port performs the feeding of biomass.

10. A multi-objective optimization device for biomass-coal co-firing, characterized in that, The step of implementing the multi-objective optimization method for biomass-coal blending according to any one of claims 1 to 9, wherein the multi-objective optimization device for biomass-coal blending comprises: The information matrix normalization module is used to acquire biomass information and coal information, and normalize and quantize them into an information matrix, wherein the information matrix includes a first information matrix of biomass information and a second information matrix of coal information. The parallel construction module is used to define multiple evaluation dimensions and decompose the information matrix, perform parallel construction and multi-objective fitting, and determine the target optimization module. The parallel construction includes a first evaluation model based on the low carbon nature of biomass-coal co-firing, a second evaluation model based on the safety of biomass-coal co-firing, and a third evaluation model based on the production ratio of biomass-coal co-firing. The full-cycle evaluation module is used to perform dimensional evaluation and integration based on the target optimization module for the entire co-firing cycle, and determine the blending ratio data. The entire co-firing cycle from the initial stage of co-firing to complete combustion is defined as a multi-stage based on the co-firing intensity.

Images