Boiler operation simulation method and system based on multi-model coupling

Abstract

The invention discloses a boiler operation simulation method and system based on multi-model coupling. The method comprises the steps that a boiler high-precision mechanism model is established based on the specific type and working condition of a boiler; establishing a CFD-AI model of the combustion process by combining a CFD technology with an AI algorithm; constructing a data driving model according to the operation data and jointly constructing a boiler dynamic model; establishing a controller based on the boiler dynamic model to perform multivariable coupling control; the system comprises a model establishment module, a coupling control module, a key influence factor identification module and a fault judgment module. The method can accurately describe the flowing, combustion and heat exchange processes in the boiler, can consider a plurality of influence factors, provides real dynamic characteristics and comprehensive analysis of the boiler system, realizes early warning and fault diagnosis, improves the operation efficiency and safety of the boiler, reduces the downtime, improves the heat efficiency of the boiler, and reduces the energy consumption of the boiler. Energy consumption and emission are reduced, support is provided for sustainable development, and cost and time in the testing and debugging process are reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of simulation management, and in particular relates to a boiler operation simulation method and system based on multi-model coupling. Background Art

[0002] As an important thermal energy production equipment, boilers are widely used in the industrial and energy fields. Their operating efficiency and environmental impact are directly related to the effective utilization and sustainable development of energy. In recent years, with the increasing global demand for energy and the gradual enhancement of environmental protection awareness, the efficient and stable operation of boiler systems is particularly important. As an important thermal energy production equipment, boilers are widely used in many fields such as industry, power generation and heating. Their performance not only affects production efficiency, but also has a profound impact on energy consumption and environmental emissions. However, traditional boiler control methods often rely on experience and simple control algorithms, which are difficult to adapt to complex operating conditions and multiple disturbance factors, resulting in low energy efficiency and excessive emissions; therefore, it is particularly important to invent a boiler operation simulation method based on multi-model coupling.

[0003] After searching, Chinese patent number CN118313282A discloses a boiler operation simulation method based on complex calculations. Although this invention improves the accuracy of the genetic algorithm and accurately obtains the best individual, so that the boiler operation simulation can be accurately carried out, it cannot accurately describe the flow, combustion and heat exchange process in the boiler, and the influencing factors are not fully considered. It cannot provide real dynamic characteristics and a comprehensive analysis of the boiler system, which reduces the operating efficiency and safety of the boiler. Summary of the invention

[0004] The purpose of the present invention is to solve the defects existing in the prior art and provide a boiler operation simulation method and system based on multi-model coupling.

[0005] In order to achieve the above object, the present invention adopts the following technical solution: The boiler operation simulation method based on multi-model coupling includes: S1. Establish a high-precision boiler mechanism model based on the specific type and operating conditions of the boiler; S2. Based on the established high-precision boiler mechanism model, a CFD-AI model of the combustion process is established using CFD technology combined with AI algorithms; S3, based on the high-precision boiler mechanism model and CFD-AI model, and based on the operation data, a data-driven model is constructed and a boiler dynamic model is jointly constructed; S4. Establish a controller based on the boiler dynamic model to perform multivariable coupling control; S5. Simulate the system response under different working conditions and identify key influencing factors through multivariable coupling control; S6. Based on the key influencing factors, judge potential faults to provide corresponding early warnings and optimize boiler operating parameters.

[0006] A further improvement of the present invention is that, based on the established high-precision boiler mechanism model in S2, a CFD-AI model of the combustion process is established by using CFD technology combined with an AI algorithm, including: Q1.1: Based on the established high-precision boiler mechanism model, obtain the corresponding geometric model inside the boiler, use the meshing tool to divide the geometric model into multiple discrete grid units, and encrypt the grid in the combustion area and airflow turbulence area; Q1.2: Obtain the flow properties in the boiler of each grid unit through each set of sensor data, including steady or unsteady, compressible or incompressible, and laminar or turbulent. Then, establish the continuity equation for the mass conservation of the fluid, and establish the momentum conservation equation by combining the conservation of mass and momentum. Establish the combustion reaction kinetic equation based on the finite rate reaction model. Q1.3: Based on the combustion reaction kinetics equation, the turbulent area generated by the combustion process is collected in real time, and the corresponding k-ε turbulence model is constructed. The k-ε turbulence model is then combined with each set of equations to construct the corresponding CFD model. After that, boundary conditions and initial conditions are set for the CFD model, and CFD simulation is performed. The control equations of the CFD model are solved iteratively using ANSYS Fluent software. Q1.4: Collect each set of data generated by CFD simulation, remove noise and outliers from each set of data, scale each set of data to the same range, use correlation analysis to evaluate feature importance, determine the key features in each set of data, and then define the number of neurons in the input layer, hidden layer, and output layer of the AI ​​model; Q1.5: Divide the extracted key features into training set, validation set and test set, use the training set data to train the model, adjust the model parameters through the back propagation algorithm, evaluate the model performance through k-fold cross validation, adjust the original hyperparameters of the AI ​​model, use the validation set to evaluate the generalization ability of the model, and calculate the accuracy and F1 score indicators; Q1.6: Repeat the training and verification of the AI ​​model until the model loss function value converges to the preset range or the accuracy and F1 score indicators meet the preset standards. Combine the prediction results of multiple groups of models through the voting method, and then verify the performance of the final model through the test set. If the model performance meets the preset indicators, the model will be deployed in actual applications, and the AI ​​model and CFD model will be coupled to form a CFD-AI combustion model.

[0007] A further improvement of the present invention is that, in S3, based on the boiler high-precision mechanism model and CFD-AI model, a data-driven model is constructed according to the operation data and a boiler dynamic model is jointly constructed, including: Q2.1: Collect the constructed high-precision boiler mechanism model, CFD-AI model and data-driven model parameter information, build the corresponding parameter space based on the key parameters of each group of models and their value ranges, and then generate multiple parameter combinations through a random number generator based on the predefined parameter range; Q2.2: Take the current state and parameter combination of the current boiler as the root node, calculate the boiler state corresponding to each randomly generated parameter combination, and take each parameter combination and the corresponding state as a new child node, and build a search tree based on the root node and child nodes; Q2.3: Starting from the root node, select child nodes along the path of the current search tree through a greedy strategy until reaching an incompletely expanded node. When reaching an incompletely expanded node, perturb the parameters of the node through a random generator and generate one or more child nodes from the node, representing new state and parameter combinations; Q2.4: Perform the simulation by randomly selecting actions or parameter combinations until the termination state is reached, and record the reward value and performance index obtained from the simulation, feed back the simulation results to all nodes on the same path, and update the reward value of each node; Q2.5: Repeat the selection, expansion, simulation, and backtracking steps until the set computing resource limit is reached. Starting from the root node, select the child node with the highest reward value as the best parameter combination, and integrate the output results of different models under this parameter combination to form a boiler dynamic model.

[0008] A further improvement of the present invention is that in S4, a controller is established based on the boiler dynamic model to perform multivariable coupling control, including: Q3.1: Based on the boiler dynamic model, obtain the variables of boiler temperature, pressure and flow that need to be controlled, as well as the relevant system status, and set the target value based on the staff's needs. Then, establish a state transition model based on the dynamic characteristics of the boiler, and associate the system status with the observation data to build the corresponding observation model; Q3.2: Select the PID closed-loop control strategy, monitor the error between the system output and the set value in real time, and then calculate the prior probability distribution of multiple variables and states based on the prior knowledge of the boiler system. Then collect the observation data generated by the observation model and update the posterior probability through Bayes' theorem; Q3.3: Based on the posterior probability, calculate the cost of the boiler state and control input, and adjust each group of control variables according to the calculated cost. Based on the optimized control variables, design a multivariable coupling control strategy and implement the control strategy in the boiler system. Meanwhile, monitor the system response in real time and continuously adjust the control input and state estimation based on the feedback data.

[0009] A further improvement of the present invention is that in S5, multivariable coupling control is used to simulate the system response under different working conditions and identify key influencing factors, including: Q4.1: Collect the range of simulatable operating conditions and set different operating condition combinations according to the operating conditions of the boiler. Use the constructed boiler dynamic model to simulate different operating conditions and output the dynamic characteristics and responses of the boiler under different operating condition combinations in real time. Q4.2: Record the changes of each input parameter and its corresponding output results, pre-process and organize the input and output data in the simulation experiment to form a data set, perform sensitivity analysis on the collected data through variance analysis, and evaluate the impact of different input parameters on the output response; Q4.3: Based on the simulation data, a linear regression model of the relationship between each group of inputs and outputs is established, and the regression coefficients of the regression model are collected. The collected coefficients are arranged from large to small. The larger the absolute value of the coefficient, the more significant the impact of the factor on the output. Based on the arrangement results, the key factors that have the greatest impact on the system response are identified, and then the identified key influencing factors are analyzed to verify their actual impact on the boiler system.

[0010] A further improvement of the present invention is that, based on the key influencing factors in S6, potential faults are judged to provide corresponding warnings and optimize boiler operating parameters, including: Q5.1: Based on the optimization goals of improving boiler thermal efficiency, reducing emissions, and reducing fuel consumption, design an optimization objective function, randomly generate multiple groups of particles, each group of particles represents a parameter combination, and randomly set the initial position and speed of each particle; Q5.2: Use the set optimization objective function to calculate the fitness value of each group of particles, initialize the global optimal position as the initial position of the first particle, update the particle position according to the current particle speed, current speed, historical optimal position and global optimal position, record the fitness value of each group of particles in each iteration, and compare the current fitness and the historical optimal fitness. If the current fitness is better than its historical optimal fitness, update the historical optimal position; Q5.3: Compare the fitness values ​​of the remaining particles with the current global optimal fitness value. If the fitness value of the current particle is better than the global optimal fitness value, update the global optimal position and the global optimal fitness value. Then, update the speed of each group of particles according to the current speed, the historical optimal position and the global optimal position, and update the position of the particles according to the updated speed. Q5.4: Check whether the updated particle position is within the preset boundary range. If it is beyond the range, adjust its position to the boundary value, and repeatedly update the particle speed, historical optimal position and global optimal position until the fitness change value reaches the preset range, and then stop updating; Q5.5: Extract the final global optimal position and the corresponding fitness value, apply the optimized control parameters to the boiler dynamic model, conduct simulation tests, and calculate and evaluate the boiler's thermal efficiency, emission level, and fuel consumption key performance indicators based on the simulation results. Compare the optimized operating parameters with the unoptimized operating parameters, analyze the difference in boiler performance before and after optimization, and evaluate the impact of each optimization parameter on boiler performance.

[0011] The boiler operation simulation system based on multi-model coupling includes: The first model building module builds a high-precision boiler mechanism model based on the specific type and operating conditions of the boiler; The second model building module uses CFD technology combined with AI algorithm to build a CFD-AI model of the combustion process based on the established high-precision boiler mechanism model; The third model building module is based on the high-precision boiler mechanism model and CFD-AI model, and builds a data-driven model based on the operation data and jointly builds the boiler dynamic model; The coupling control module establishes a controller based on the boiler dynamic model to perform multivariable coupling control; Identify key influencing factors module, simulate the system response under different working conditions and identify key influencing factors through multivariable coupling control; The fault judgment module judges potential faults based on key influencing factors to provide corresponding early warnings and optimize boiler operating parameters.

[0012] A further improvement of the present invention is that, in the second model building module, based on the established high-precision boiler mechanism model, a CFD-AI model of the combustion process is established by using CFD technology combined with an AI algorithm, including: Q1.1: Based on the established high-precision boiler mechanism model, obtain the corresponding geometric model inside the boiler, use the meshing tool to divide the geometric model into multiple discrete grid units, and encrypt the grid in the combustion area and airflow turbulence area; Q1.2: Obtain the flow properties in the boiler of each grid unit through each set of sensor data, including steady or unsteady, compressible or incompressible, and laminar or turbulent. Then, establish the continuity equation for the mass conservation of the fluid, and establish the momentum conservation equation by combining the conservation of mass and momentum. Establish the combustion reaction kinetic equation based on the finite rate reaction model. Q1.3: Based on the combustion reaction kinetics equation, the turbulent area generated by the combustion process is collected in real time, and the corresponding k-ε turbulence model is constructed. The k-ε turbulence model is then combined with each set of equations to construct the corresponding CFD model. After that, boundary conditions and initial conditions are set for the CFD model, and CFD simulation is performed. The control equations of the CFD model are solved iteratively using ANSYS Fluent software. Q1.4: Collect each set of data generated by CFD simulation, remove noise and outliers from each set of data, scale each set of data to the same range, use correlation analysis to evaluate feature importance, determine the key features in each set of data, and then define the number of neurons in the input layer, hidden layer, and output layer of the AI ​​model; Q1.5: Divide the extracted key features into training set, validation set and test set, use the training set data to train the model, adjust the model parameters through the back propagation algorithm, evaluate the model performance through k-fold cross validation, adjust the original hyperparameters of the AI ​​model, use the validation set to evaluate the generalization ability of the model, and calculate the accuracy and F1 score indicators; Q1.6: Repeat the training and verification of the AI ​​model until the model loss function value converges to the preset range or the accuracy and F1 score indicators meet the preset standards. Combine the prediction results of multiple groups of models through the voting method, and then verify the performance of the final model through the test set. If the model performance meets the preset indicators, the model will be deployed in actual applications, and the AI ​​model and CFD model will be coupled to form a CFD-AI combustion model.

[0013] A further improvement of the present invention is that the third model building module is based on the boiler high-precision mechanism model and CFD-AI model, and a data-driven model is built according to the operation data and a boiler dynamic model is jointly built, including: Q2.1: Collect the constructed high-precision boiler mechanism model, CFD-AI model and data-driven model parameter information, build the corresponding parameter space based on the key parameters of each group of models and their value ranges, and then generate multiple parameter combinations through a random number generator based on the predefined parameter range; Q2.2: Take the current state and parameter combination of the current boiler as the root node, calculate the boiler state corresponding to each randomly generated parameter combination, and take each parameter combination and the corresponding state as a new child node, and build a search tree based on the root node and child nodes; Q2.3: Starting from the root node, select child nodes along the path of the current search tree through a greedy strategy until reaching an incompletely expanded node. When reaching an incompletely expanded node, perturb the parameters of the node through a random generator and generate one or more child nodes from the node, representing new state and parameter combinations; Q2.4: Perform the simulation by randomly selecting actions or parameter combinations until the termination state is reached, and record the reward value and performance index obtained from the simulation, feed back the simulation results to all nodes on the same path, and update the reward value of each node; Q2.5: Repeat the selection, expansion, simulation, and backtracking steps until the set computing resource limit is reached. Starting from the root node, select the child node with the highest reward value as the best parameter combination, and integrate the output results of different models under this parameter combination to form a boiler dynamic model.

[0014] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the boiler operation simulation method based on multi-model coupling are implemented.

[0015] Compared with the prior art, the present invention has at least the following beneficial technical effects: The boiler operation simulation method and system based on multi-model coupling provided by the present invention build a high-precision mechanism model based on physical laws, combine CFD and AI algorithms, build a CFD-AI combustion model to simulate the combustion process, and use machine learning methods to learn the dynamic characteristics of the boiler from historical data to build a data-driven model. Then, a corresponding parameter space is built based on the key parameters of each group of models and their value ranges. Then, according to the predefined parameter range, multiple parameter combinations are generated through a random number generator, and a search tree is built to select the optimal parameter combination. The output results of different models under the parameter combination are integrated to form a comprehensive boiler dynamic model. Based on the boiler dynamic model, the groups of variables of boiler temperature, pressure and flow that need to be controlled, as well as related system states, are obtained, and target values ​​are set based on staff needs. Then, a state transition model is established based on the dynamic characteristics of the boiler, and the system state is associated with the observed data to build a corresponding observation model. Select PID closed-loop control strategy monitors the error between system output and set value in real time, randomly generates multiple groups of particles, each group of particles represents a parameter combination, and randomly sets the initial position and speed of each particle. The set optimization objective function is used to calculate the fitness value of each group of particles, and the position, speed, historical optimal position and global optimal position of each group of particles are continuously updated and iterated. The optimized operating parameters are compared with the operating parameters before and after optimization, and the difference in boiler performance before and after optimization is analyzed. The influence of each optimization parameter on boiler performance is evaluated, which can accurately describe the flow, combustion and heat exchange process in the boiler, and can consider multiple influencing factors at the same time, provide real dynamic characteristics and comprehensive analysis of the boiler system, realize early warning and fault diagnosis, improve the operating efficiency and safety of the boiler, reduce downtime, improve the thermal efficiency of the boiler, reduce energy consumption and emissions, provide support for sustainable development, and reduce the cost and time during testing and debugging. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0017] Figure 1 The present invention is a flow chart of a boiler operation simulation method based on multi-model coupling.

[0018] Figure 2 It is a structural block diagram of the boiler operation simulation system based on multi-model coupling of the present invention. DETAILED DESCRIPTION

[0019] In the following, only some exemplary embodiments are briefly described. As those skilled in the art will appreciate, the described embodiments may be modified in various ways without departing from the spirit or scope of the present invention. Therefore, the drawings and descriptions are considered to be exemplary and non-restrictive in nature.

[0020] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0021] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0022] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0023] Various structural schematic diagrams of the embodiments disclosed in the present invention are shown in the accompanying drawings. These figures are not drawn to scale, and some details are magnified and some details may be omitted for the purpose of clear expression. The shapes of various regions and layers shown in the figures and the relative sizes and positional relationships therebetween are only exemplary, and may deviate in practice due to manufacturing tolerances or technical limitations, and those skilled in the art may additionally design regions / layers with different shapes, sizes, and relative positions according to actual needs.

[0024] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0025] Example 1 The boiler operation simulation method based on multi-model coupling provided by the present invention includes: S1. Establish a high-precision boiler mechanism model based on the specific type and operating conditions of the boiler; S2. Based on the established high-precision boiler mechanism model, a CFD-AI model of the combustion process is established using CFD technology combined with AI algorithms; S3, based on the high-precision boiler mechanism model and CFD-AI model, and based on the operation data, a data-driven model is constructed and a boiler dynamic model is jointly constructed; S4. Establish a controller based on the boiler dynamic model to perform multivariable coupling control; S5. Simulate the system response under different working conditions and identify key influencing factors through multivariable coupling control; S6. Based on the key influencing factors, judge potential faults to provide corresponding early warnings and optimize boiler operating parameters.

[0026] Example 2 Reference Figure 1 The boiler operation simulation method based on multi-model coupling provided by the present invention has the following specific steps: A high-precision boiler mechanism model is established based on the specific type and operating conditions of the boiler.

[0027] Based on the established high-precision boiler mechanism model, the CFD technology is combined with the AI ​​algorithm to establish a CFD-AI model of the combustion process.

[0028] Specifically, based on the established high-precision boiler mechanism model, the geometric model of the corresponding boiler interior is obtained, and the meshing tool is used to divide the geometric model into multiple discrete mesh units, and the mesh is encrypted in the combustion area and the airflow turbulence area. The properties of the flow in each mesh unit of the boiler are obtained through each group of sensor data, including steady state or unsteady state, compressible or incompressible, and laminar or turbulent flow. Then, for the conservation of mass of the fluid, a continuity equation is established, and in combination with the conservation of mass and momentum, a momentum conservation equation is established. According to the finite rate reaction model, a combustion reaction kinetic equation is established. Based on the combustion reaction kinetic equation, the turbulent area generated by the combustion process is collected in real time, and the corresponding k-ε turbulence model is constructed. The k-ε turbulence model is then combined with each group of equations to construct the corresponding CFD model. Then, boundary conditions and initial conditions are set for the CFD model, and CFD simulation is performed. Through ANSYS Fluent software solves the control equations of the CFD model in an iterative manner, collects each group of data generated by the CFD simulation, removes the noise and outliers of each group of data, and then scales each group of processed data to the same range. Correlation analysis is used to evaluate the importance of features to determine the key features in each group of data, and then the number of neurons in the input layer, hidden layer, and output layer of the AI ​​model is defined. The extracted key features are divided into training set, validation set, and test set. The training set data is used for model training, and the model parameters are adjusted through the back propagation algorithm. The model performance is evaluated through k-fold cross validation, and the original hyperparameters of the AI ​​model are adjusted. The generalization ability of the model is evaluated using the validation set, and the accuracy and F1 score indicators are calculated. The training and verification of the AI ​​model are repeated until the model loss function value converges to the preset range or the accuracy and F1 score indicators meet the preset standards. The prediction results of multiple groups of models are combined through the voting method, and then the performance of the final model is verified through the test set. If the model performance meets the preset indicators, the model is deployed in actual applications, and the AI ​​model and CFD model are coupled to form a CFD-AI combustion model.

[0029] Based on the high-precision mechanism model and CFD-AI model of the boiler, a data-driven model is constructed based on the operating data and the boiler dynamic model is jointly constructed.

[0030] Specifically, the constructed high-precision boiler mechanism model, CFD-AI model and data-driven model parameter information are collected, and the corresponding parameter space is constructed based on the key parameters of each group of models and their value ranges. Then, according to the predefined parameter range, multiple parameter combinations are generated through a random number generator. The current state and parameter combination of the current boiler are used as the root node, and the boiler state corresponding to each randomly generated group of parameter combinations is calculated. Each group of parameter combinations and the corresponding state are used as new child nodes. A search tree is constructed based on the root node and child nodes. Starting from the root node, the child nodes are selected along the path of the current search tree through a greedy strategy until a node that is not fully expanded is reached. When it is reached, When reaching a node that is not fully expanded, the parameters of the node are perturbed through the random generator, and one or more child nodes are generated from the node to represent the new state and parameter combination. The process is carried out by randomly selecting actions or parameter combinations until the termination state is reached, and the reward value and performance index obtained by the simulation are recorded. The simulation results are fed back to all nodes on the same path, and the reward value of each node is updated. The selection, expansion, simulation and backtracking steps are repeated until the set computing resource limit is reached. Starting from the root node, the child node with the highest reward value is selected as the best parameter combination, and the output results of different models under this parameter combination are integrated to form a boiler dynamic model.

[0031] Example 3 Reference Figure 1 , a boiler operation simulation method based on multi-model coupling, the specific steps of the simulation method are as follows: A controller is established based on the boiler dynamic model to perform multivariable coupling control.

[0032] Specifically, based on the boiler dynamic model, the various groups of variables of boiler temperature, pressure and flow that need to be controlled, as well as the relevant system status, are obtained, and the target value is set based on the needs of the staff. Then, a state transition model is established according to the dynamic characteristics of the boiler, and the system status is associated with the observation data to construct the corresponding observation model, select the PID closed-loop control strategy, and monitor the error between the system output and the set value in real time. Then, based on the prior knowledge of the boiler system, the prior probability distribution of multiple variables and states is calculated, and then the various groups of observation data generated by the observation model are collected, and the posterior probability is updated through the Bayesian theorem. Based on the posterior probability, the cost of the boiler state and control input is calculated, and according to the calculated cost, each group of control variables is adjusted. Based on the optimized control variables, a multivariable coupling control strategy is designed, and the control strategy is implemented in the boiler system. At the same time, the system response is monitored in real time, and the control input and state estimation are continuously adjusted according to the feedback data.

[0033] Multivariable coupling control is used to simulate system responses under different operating conditions and identify key influencing factors.

[0034] Specifically, collect the range of simulatable operating conditions, and set different operating condition combinations according to the operating conditions of the boiler. Use the constructed boiler dynamic model to simulate under different operating conditions, and output the dynamic characteristics and responses of the boiler under different operating condition combinations in real time. Record the changes in each input parameter and its corresponding output results. Preprocess and organize the input and output data in the simulation experiment to form a data set. Perform sensitivity analysis on the collected data through variance analysis to evaluate the influence of different input parameters on the output response. Establish a linear regression model of the relationship between each group of input and output based on the simulation data, and collect the regression coefficients of the regression model. Arrange the collected coefficients from large to small. The larger the absolute value of the coefficient, the more significant the influence of the factor on the output. According to the arrangement results, identify the key factors that have the greatest influence on the system response, and then analyze the identified key influencing factors to verify their actual influence on the boiler system.

[0035] Based on the key influencing factors, potential faults are judged to provide corresponding early warnings and optimize boiler operating parameters.

[0036] Specifically, based on the optimization objectives of improving the thermal efficiency of the boiler, reducing emissions and reducing fuel consumption, an optimization objective function is designed, multiple groups of particles are randomly generated, each group of particles represents a parameter combination, and the initial position and speed of each particle are randomly set. The fitness value of each group of particles is calculated using the set optimization objective function, and the global optimal position is initialized as the initial position of the first particle. The particle position is updated according to the current particle speed, the current speed, the historical optimal position and the global optimal position. The fitness value of each group of particles in each iteration is recorded, and the current fitness and the historical optimal fitness are compared. If the current fitness is better than its historical optimal fitness, the historical optimal position is updated, and the relationship between the fitness values ​​of the remaining particles and the current global optimal fitness value is compared. If the fitness value of the current particle is better than the global optimal fitness value, the global optimal position is updated, and the global optimal fitness value is updated at the same time. The fitness value is obtained, and then the speed of each group of particles is updated according to the current speed, the historical optimal position and the global optimal position, and the position of the particles is updated according to the updated speed. It is checked whether the updated particle position is within the preset boundary range. If it is out of the range, its position is adjusted to the boundary value. The particle speed, the historical optimal position and the global optimal position are repeatedly updated until the fitness change value reaches the preset range, and the updating is stopped. The final global optimal position and the corresponding fitness value are extracted, and the optimized control parameters are applied to the boiler dynamic model for simulation testing. According to the simulation results, the thermal efficiency, emission level and fuel consumption of the boiler are calculated and evaluated. The optimized operating parameters are compared with the operating parameters before optimization, the difference in boiler performance before and after optimization is analyzed, and the influence of each optimization parameter on boiler performance is evaluated.

[0037] Example 4 Reference Figure 2 The boiler operation simulation system based on multi-model coupling provided by the present invention comprises: The first model building module builds a high-precision boiler mechanism model based on the specific type and operating conditions of the boiler; The second model building module uses CFD technology combined with AI algorithm to build a CFD-AI model of the combustion process based on the established high-precision boiler mechanism model; The third model building module is based on the high-precision boiler mechanism model and CFD-AI model, and builds a data-driven model based on the operation data and jointly builds the boiler dynamic model; The coupling control module establishes a controller based on the boiler dynamic model to perform multivariable coupling control; Identify key influencing factors module, simulate the system response under different working conditions and identify key influencing factors through multivariable coupling control; The fault judgment module judges potential faults based on key influencing factors to provide corresponding early warnings and optimize boiler operating parameters.

[0038] Example 5 The present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the boiler operation simulation method based on multi-model coupling are implemented.

[0039] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0040] The present application is described with reference to the flowcharts and / or block diagrams of the methods, systems and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A system that specifies the functions of a box or multiple boxes.

[0041] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0042] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0043] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention. Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the attached claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any figure mark in the claims should not be regarded as limiting the claims involved.

[0044] In addition, it should be understood that although this specification is described in accordance with the implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation modes that can be understood by those skilled in the art. The above content is only to illustrate the technical idea of ​​the present invention, and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution according to the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A boiler operation simulation method based on multi-model coupling, characterized in that: include: S1. Establish a high-precision boiler mechanism model based on the specific type and operating conditions of the boiler; S2. Based on the established high-precision boiler mechanism model, a CFD-AI model of the combustion process is established using CFD technology combined with AI algorithms; S3, based on the high-precision boiler mechanism model and CFD-AI model, and based on the operation data, a data-driven model is constructed and a boiler dynamic model is jointly constructed; S4. Establish a controller based on the boiler dynamic model to perform multivariable coupling control; S5. Simulate the system response under different working conditions and identify key influencing factors through multivariable coupling control; S6. Based on the key influencing factors, judge potential faults to provide corresponding early warnings and optimize boiler operating parameters.

2. The boiler operation simulation method based on multi-model coupling according to claim 1 is characterized in that: In S2, based on the established high-precision boiler mechanism model, CFD technology combined with AI algorithm is used to establish a CFD-AI model of the combustion process, including: Q1.1: Based on the established high-precision boiler mechanism model, obtain the corresponding geometric model inside the boiler, use the meshing tool to divide the geometric model into multiple discrete grid units, and encrypt the grid in the combustion area and airflow turbulence area; Q1.2: Obtain the flow properties in the boiler of each grid unit through each set of sensor data, including steady or unsteady, compressible or incompressible, and laminar or turbulent. Then, establish the continuity equation for the mass conservation of the fluid, and establish the momentum conservation equation by combining the conservation of mass and momentum. Establish the combustion reaction kinetic equation based on the finite rate reaction model. Q1.3: Based on the combustion reaction kinetics equation, the turbulent area generated by the combustion process is collected in real time, and the corresponding k-ε turbulence model is constructed. The k-ε turbulence model is then combined with each set of equations to construct the corresponding CFD model. After that, boundary conditions and initial conditions are set for the CFD model, and CFD simulation is performed. The control equations of the CFD model are solved iteratively using ANSYS Fluent software. Q1.4: Collect each set of data generated by CFD simulation, remove noise and outliers from each set of data, scale each set of data to the same range, use correlation analysis to evaluate feature importance, determine the key features in each set of data, and then define the number of neurons in the input layer, hidden layer, and output layer of the AI ​​model; Q1.5: Divide the extracted key features into training set, validation set and test set, use the training set data to train the model, adjust the model parameters through the back propagation algorithm, evaluate the model performance through k-fold cross validation, adjust the original hyperparameters of the AI ​​model, use the validation set to evaluate the generalization ability of the model, and calculate the accuracy and F1 score indicators; Q1.6: Repeat the training and verification of the AI ​​model until the model loss function value converges to the preset range or the accuracy and F1 score indicators meet the preset standards. Combine the prediction results of multiple groups of models through the voting method, and then verify the performance of the final model through the test set. If the model performance meets the preset indicators, the model will be deployed in actual applications, and the AI ​​model and CFD model will be coupled to form a CFD-AI combustion model.

3. The boiler operation simulation method based on multi-model coupling according to claim 2 is characterized in that: S3 builds a data-driven model based on the high-precision boiler mechanism model and CFD-AI model, and jointly builds a boiler dynamic model based on operating data, including: Q2.1: Collect the constructed high-precision boiler mechanism model, CFD-AI model and data-driven model parameter information, build the corresponding parameter space based on the key parameters of each group of models and their value ranges, and then generate multiple parameter combinations through a random number generator based on the predefined parameter range; Q2.2: Take the current state and parameter combination of the current boiler as the root node, calculate the boiler state corresponding to each randomly generated parameter combination, and take each parameter combination and the corresponding state as a new child node, and build a search tree based on the root node and child nodes; Q2.3: Starting from the root node, select child nodes along the path of the current search tree through a greedy strategy until you reach a node that is not fully expanded. When you reach a node that is not fully expanded, perturb the parameters of the node through a random generator and generate one or more child nodes from the node, representing new state and parameter combinations; Q2.4: Perform the simulation by randomly selecting actions or parameter combinations until the termination state is reached, and record the reward value and performance index obtained from the simulation, feed back the simulation results to all nodes on the same path, and update the reward value of each node; Q2.5: Repeat the selection, expansion, simulation, and backtracking steps until the set computing resource limit is reached. Starting from the root node, select the child node with the highest reward value as the best parameter combination, and integrate the output results of different models under this parameter combination to form a boiler dynamic model.

4. The boiler operation simulation method based on multi-model coupling according to claim 3 is characterized in that: In S4, a controller is established based on the boiler dynamic model to perform multivariable coupling control, including: Q3.1: Based on the boiler dynamic model, obtain the variables of boiler temperature, pressure and flow that need to be controlled, as well as the relevant system status, and set the target value based on the staff's needs. Then, establish a state transition model based on the dynamic characteristics of the boiler, and associate the system status with the observation data to build the corresponding observation model; Q3.2: Select the PID closed-loop control strategy, monitor the error between the system output and the set value in real time, and then calculate the prior probability distribution of multiple variables and states based on the prior knowledge of the boiler system. Then collect the observation data generated by the observation model and update the posterior probability through Bayes' theorem; Q3.3: Based on the posterior probability, calculate the cost of the boiler state and control input, and adjust each group of control variables according to the calculated cost. Based on the optimized control variables, design a multivariable coupling control strategy and implement the control strategy in the boiler system. Meanwhile, monitor the system response in real time and continuously adjust the control input and state estimation based on the feedback data.

5. The boiler operation simulation method based on multi-model coupling according to claim 3 is characterized in that: In S5, multivariable coupling control is used to simulate the system response under different working conditions and identify key influencing factors, including: Q4.1: Collect the range of simulatable operating conditions and set different operating condition combinations according to the operating conditions of the boiler. Use the constructed boiler dynamic model to simulate different operating conditions and output the dynamic characteristics and responses of the boiler under different operating condition combinations in real time. Q4.2: Record the changes of each input parameter and its corresponding output results, pre-process and organize the input and output data in the simulation experiment to form a data set, perform sensitivity analysis on the collected data through variance analysis, and evaluate the impact of different input parameters on the output response; Q4.3: Based on the simulation data, a linear regression model of the relationship between each group of inputs and outputs is established, and the regression coefficients of the regression model are collected. The collected coefficients are arranged from large to small. The larger the absolute value of the coefficient, the more significant the impact of the factor on the output. Based on the arrangement results, the key factors that have the greatest impact on the system response are identified, and then the identified key influencing factors are analyzed to verify their actual impact on the boiler system.

6. The boiler operation simulation method based on multi-model coupling according to claim 5 is characterized in that: S6 uses key influencing factors to identify potential faults, provide corresponding warnings and optimize boiler operating parameters, including: Q5.1: Based on the optimization goals of improving boiler thermal efficiency, reducing emissions, and reducing fuel consumption, design an optimization objective function, randomly generate multiple groups of particles, each group of particles represents a parameter combination, and randomly set the initial position and speed of each particle; Q5.2: Use the set optimization objective function to calculate the fitness value of each group of particles, initialize the global optimal position as the initial position of the first particle, update the particle position according to the current particle speed, current speed, historical optimal position and global optimal position, record the fitness value of each group of particles in each iteration, and compare the current fitness and the historical optimal fitness. If the current fitness is better than its historical optimal fitness, update the historical optimal position; Q5.3: Compare the fitness values ​​of the remaining particles with the current global optimal fitness value. If the fitness value of the current particle is better than the global optimal fitness value, update the global optimal position and the global optimal fitness value. Then, update the speed of each group of particles according to the current speed, the historical optimal position and the global optimal position, and update the position of the particles according to the updated speed. Q5.4: Check whether the updated particle position is within the preset boundary range. If it is beyond the range, adjust its position to the boundary value, and repeatedly update the particle speed, historical optimal position and global optimal position until the fitness change value reaches the preset range, and then stop updating; Q5.5: Extract the final global optimal position and the corresponding fitness value, apply the optimized control parameters to the boiler dynamic model, conduct simulation tests, and calculate and evaluate the boiler's thermal efficiency, emission level, and fuel consumption key performance indicators based on the simulation results. Compare the optimized operating parameters with the unoptimized operating parameters, analyze the difference in boiler performance before and after optimization, and evaluate the impact of each optimization parameter on boiler performance.

7. The boiler operation simulation system based on multi-model coupling is characterized by: include: The first model building module builds a high-precision boiler mechanism model based on the specific type and operating conditions of the boiler; The second model building module uses CFD technology combined with AI algorithm to build a CFD-AI model of the combustion process based on the established high-precision boiler mechanism model; The third model building module is based on the high-precision boiler mechanism model and CFD-AI model, and builds a data-driven model based on the operation data and jointly builds the boiler dynamic model; The coupling control module establishes a controller based on the boiler dynamic model to perform multivariable coupling control; Identify key influencing factors module, simulate the system response under different working conditions and identify key influencing factors through multivariable coupling control; The fault judgment module judges potential faults based on key influencing factors to provide corresponding early warnings and optimize boiler operating parameters.

8. The boiler operation simulation system based on multi-model coupling according to claim 7 is characterized in that: In the second model building module, based on the established high-precision boiler mechanism model, the CFD technology is combined with the AI ​​algorithm to establish the CFD-AI model of the combustion process, including: Q1.1: Based on the established high-precision boiler mechanism model, obtain the corresponding geometric model inside the boiler, use the meshing tool to divide the geometric model into multiple discrete grid units, and encrypt the grid in the combustion area and airflow turbulence area; Q1.2: Obtain the flow properties in the boiler of each grid unit through each set of sensor data, including steady or unsteady, compressible or incompressible, and laminar or turbulent. Then, establish the continuity equation for the mass conservation of the fluid, and establish the momentum conservation equation by combining the conservation of mass and momentum. Establish the combustion reaction kinetic equation based on the finite rate reaction model. Q1.3: Based on the combustion reaction kinetics equation, the turbulent area generated by the combustion process is collected in real time, and the corresponding k-ε turbulence model is constructed. The k-ε turbulence model is then combined with each set of equations to construct the corresponding CFD model. After that, boundary conditions and initial conditions are set for the CFD model, and CFD simulation is performed. The control equations of the CFD model are solved iteratively using ANSYS Fluent software. Q1.4: Collect each set of data generated by CFD simulation, remove noise and outliers from each set of data, scale each set of data to the same range, use correlation analysis to evaluate feature importance, determine the key features in each set of data, and then define the number of neurons in the input layer, hidden layer, and output layer of the AI ​​model; Q1.5: Divide the extracted key features into training set, validation set and test set, use the training set data to train the model, adjust the model parameters through the back propagation algorithm, evaluate the model performance through k-fold cross validation, adjust the original hyperparameters of the AI ​​model, use the validation set to evaluate the generalization ability of the model, and calculate the accuracy and F1 score indicators; Q1.6: Repeat the training and verification of the AI ​​model until the model loss function value converges to the preset range or the accuracy and F1 score indicators meet the preset standards. Combine the prediction results of multiple groups of models through the voting method, and then verify the performance of the final model through the test set. If the model performance meets the preset indicators, the model will be deployed in actual applications, and the AI ​​model and CFD model will be coupled to form a CFD-AI combustion model.

9. The boiler operation simulation system based on multi-model coupling according to claim 8 is characterized in that: The third model building module is based on the boiler high-precision mechanism model and CFD-AI model, and builds a data-driven model based on the operation data and jointly builds a boiler dynamic model, including: Q2.1: Collect the constructed high-precision boiler mechanism model, CFD-AI model and data-driven model parameter information, build the corresponding parameter space based on the key parameters of each group of models and their value ranges, and then generate multiple parameter combinations through a random number generator based on the predefined parameter range; Q2.2: Take the current state and parameter combination of the current boiler as the root node, calculate the boiler state corresponding to each randomly generated parameter combination, and take each parameter combination and the corresponding state as a new child node, and build a search tree based on the root node and child nodes; Q2.3: Starting from the root node, select child nodes along the path of the current search tree through a greedy strategy until you reach a node that is not fully expanded. When you reach a node that is not fully expanded, perturb the parameters of the node through a random generator and generate one or more child nodes from the node, representing new state and parameter combinations; Q2.4: Perform the simulation by randomly selecting actions or parameter combinations until the termination state is reached, and record the reward value and performance index obtained from the simulation, feed back the simulation results to all nodes on the same path, and update the reward value of each node; Q2.5: Repeat the selection, expansion, simulation, and backtracking steps until the set computing resource limit is reached. Starting from the root node, select the child node with the highest reward value as the best parameter combination, and integrate the output results of different models under this parameter combination to form a boiler dynamic model.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the boiler operation simulation method based on multi-model coupling according to any one of claims 1 to 6 are implemented.

Images