Optimization and energy saving method of boiler system based on energy balance

Abstract

The present invention discloses a boiler system optimization and energy-saving method based on energy balance, which belongs to the field of energy engineering technology, and specifically includes: collecting boiler operation data in real time through sensors, constructing a three-dimensional thermal flow field model using an improved CFD method, and predicting exhaust heat loss in combination with deep learning; generating a multi-objective optimization strategy using a genetic algorithm and a particle swarm optimization algorithm, and implementing energy-saving optimization measures; establishing a closed-loop feedback mechanism, monitoring the optimization effect in real time, and dynamically adjusting the strategy using a reinforcement learning method; performing energy balance analysis, continuously monitoring and evaluating the energy-saving effect, thereby improving the energy efficiency of the boiler system and achieving the goal of energy conservation and emission reduction.

Description

Technical Field

[0001] The invention belongs to the technical field of energy engineering, and in particular to an energy-balancing-based boiler system optimization and energy-saving method. Background Art

[0002] With the continuous growth of global energy demand and the increasingly serious environmental problems, improving the energy utilization efficiency of industrial boilers and reducing energy consumption and environmental pollution have become the focus of global attention. At present, many industrial boilers have energy waste during operation. Due to aging equipment, improper operation and unreasonable system design, the thermal efficiency of boilers is generally low and cannot achieve the ideal energy utilization level. Secondly, the energy waste of industrial boilers not only increases the operating costs of enterprises, but also has adverse effects on the environment.

[0003] For example, the patent with the authorization announcement number CN103234190B discloses a chain furnace energy-saving optimization control system, including: adjusting the air volume according to the main steam pressure, adjusting the grate speed according to the flue gas oxygen content, and adjusting the coal layer thickness according to the load and coal quality. The invention is beneficial in that the timeliness of load adjustment is faster than adjusting the coal supply by the main steam pressure; the load-oxygen setter control method can achieve low flue gas oxygen content at high load and high flue gas oxygen content at low load, ensuring a reasonable air excess coefficient, which is conducive to improving the thermal efficiency of the boiler; the load increases, the coal layer thickens; the coal quality deteriorates, the coal layer thickens, making the grate speed more reasonable, avoiding the problem of too short coal residence time and incomplete coal combustion caused by high grate speed, thereby improving the thermal efficiency of the boiler.

[0004] The above existing technologies all have the following problems: 1) They rely on manual or simple feedback mechanisms to adjust the air volume, grate speed and coal seam thickness, and lack real-time data-driven models to accurately predict and adjust operating parameters; 2) They adjust parameters through fixed logic or rules, and cannot achieve optimal results in all cases; 3) They cannot make full use of large amounts of historical data and real-time data to improve control strategies. Summary of the invention

[0005] In view of the shortcomings of the existing technology, the present invention proposes a boiler system optimization and energy-saving method based on energy balance. The boiler operation data is collected in real time by sensors, and a three-dimensional thermal flow field model is constructed using an improved CFD method. The exhaust heat loss is predicted in combination with deep learning. A genetic algorithm and a particle swarm optimization algorithm are used to generate a multi-objective optimization strategy, and energy-saving optimization measures are implemented. A closed-loop feedback mechanism is established to monitor the optimization effect in real time, and a reinforcement learning method is used to dynamically adjust the strategy. Energy balance analysis is performed, and the energy-saving effect is continuously monitored and evaluated, thereby improving the energy efficiency of the boiler system and achieving the goal of energy conservation and emission reduction.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] Boiler system optimization and energy saving methods based on energy balance include:

[0008] Step S1: The operation data of the boiler system is collected in real time through sensors and preprocessed. Based on the preprocessed operation data, a three-dimensional thermal flow field model of the boiler is constructed using the CFD method, and a real-time data update mechanism is introduced to dynamically simulate the heat distribution during the exhaust process. The three-dimensional thermal flow field model of the boiler is trained by combining the deep learning method and physical rule constraints to obtain the prediction result of exhaust heat loss;

[0009] Step S2: setting a reduction target value for boiler exhaust heat loss, using a genetic algorithm combined with a particle swarm optimization algorithm to generate a multi-objective optimization strategy set, and formulating energy-saving optimization measures based on the multi-objective optimization strategy set;

[0010] Step S3: Establish a closed-loop feedback mechanism to monitor the execution effect of energy-saving optimization measures in real time, and dynamically adjust the multi-objective optimization strategy according to the execution effect;

[0011] Step S4: After the energy-saving optimization measures are implemented, the performance indicators of the boiler system are tested and evaluated, and a real-time feedback mechanism is established to evaluate the optimization effect in real time based on the monitoring data to determine whether the optimization effect has achieved the expected goal. If the expected goal has not been achieved, a genetic algorithm is used for iterative optimization, and the multi-objective optimization strategy set is updated;

[0012] Step S5: Regularly perform energy balance analysis on the boiler system to identify potential energy-saving points, and establish a data tracking mechanism to continuously monitor the operation status and energy-saving effect of the boiler system, and regularly evaluate the energy-saving effect.

[0013] Specifically, the specific steps of step S1 include:

[0014] S1.1: Use sensors to collect the operating data of the boiler system in real time and perform preprocessing to obtain the preprocessed operating data X = {x 1 ,...,x n}, where x n Indicates the nth running data, where n indicates the number of running data;

[0015] S1.2: Based on X = {x 1 ,...,x n}, using CFD method to build a three-dimensional model of the boiler according to the geometric structure of the boiler, using momentum conservation equations to obtain fluid flow and energy transfer process, combining the law of conservation of energy to establish the heat balance equation, and X = {x 1 ,...,x n} Input the three-dimensional model as boundary conditions to build a three-dimensional thermal flow field model of the boiler;

[0016] S1.3: Introduce a real-time data update mechanism to dynamically simulate the heat distribution during the exhaust process based on real-time operation data. Based on the simulation results, train the three-dimensional thermal flow field model of the boiler in combination with deep learning methods.

[0017] S1.4: Set the loss function, add the constraints of energy conservation, heat conduction equation, and momentum conservation to the loss function, and obtain the exhaust heat loss prediction results of the boiler three-dimensional thermal flow field model based on real-time operation data.

[0018] Specifically, the heat balance equation in S1.2 is:

[0019]

[0020] Among them, S los represents exhaust heat loss, S fir represents the heat released by fuel combustion, ρ represents the fluid density, u i and u j They all represent the velocity components of the fluid in the x-axis, y-axis, and z-axis directions, and i,j∈[1,3],i≠j,y i represents the spatial coordinates of the fluid in the x-axis, y-axis, and z-axis directions, τ ij represents the component of the viscous stress tensor, represents the component of the heat flux vector caused by heat conduction in the i direction, It represents the component of the heat flux vector caused by radiation in the i direction, E represents the internal energy and kinetic energy of the fluid per unit mass, t represents the current time, and h represents the pressure.

[0021] Specifically, the loss function formula after adding the constraint term in S1.4 is:

[0022]

[0023] Among them, L tol represents the loss function after adding constraints, represents the prediction result of the kth running data, λ represents the weight of the regularization term, L 1 (T,ρ) represents the loss function L 1 Regularization term, T represents temperature, L en represents the energy conservation constraint, γ 1 represents the weight of the energy conservation constraint, L he represents the constraint term of the heat conduction equation, γ 2 represents the weight of the constraint term of the heat conduction equation, L mo represents the momentum conservation constraint, γ 3 Represents the weight of the momentum conservation constraint.

[0024] Specifically, the specific steps of step S2 include:

[0025] S2.1: Set the target value for reducing exhaust heat loss, and obtain X = {x 1 ,...,x n} data, and define the decision variables and objective function, the formula is:

[0026] f(z)=S los (z 1 ,...,z l )

[0027] Where f(z) represents the objective function of minimizing the heat loss of smoke exhaust, (z 1 ,...,z l ) represents the decision variable, z l represents the lth decision variable, S los (·) represents the exhaust heat loss function;

[0028] S2.2: Based on the three-dimensional thermal flow field model of the boiler established in S1.2 and the physical rule constraints in S1.4, construct the constraints of the optimization problem. The constraints include:

[0029]

[0030] Among them, p 1 represents boiler input energy, p 2 represents boiler output energy, β represents thermal conductivity, ▽T represents temperature gradient, T wal represents the wall temperature, T wal,sp Indicates the maximum temperature that the wall can withstand, F fu Indicates the fuel flow rate, F fu,min Indicates the minimum fuel flow rate, F fu,max Indicates the maximum fuel flow rate.

[0031] Specifically, the specific steps of step S2 also include:

[0032] S2.3: Generate a set of multi-objective optimization strategies using a hybrid strategy of genetic algorithm and particle swarm optimization algorithm, and evaluate the generated multi-objective optimization strategies using the Pareto frontier method;

[0033] S2.4: Determine whether the evaluation result meets the multi-objective optimization conditions. If not, iteratively optimize the multi-objective optimization strategy to obtain the optimal multi-objective optimization strategy;

[0034] S2.5: Develop energy-saving optimization measures based on the optimal multi-objective optimization strategy.

[0035] Specifically, the specific steps of S2.3 include:

[0036] S2.31: Randomly initialize the position and speed of the particle swarm, set the size of the particle swarm, the number of iterations, the crossover and mutation operation parameters of the genetic algorithm, and set the current position of each particle as its individual optimal solution;

[0037] S2.32: For each candidate strategy in the genetic algorithm and the particle swarm optimization algorithm, a three-dimensional thermal flow field model of the boiler is used for simulation, and the heat loss corresponding to each candidate strategy is extracted from the simulation results, and the extracted heat loss data is input into f(z) to obtain the objective function value of each candidate strategy;

[0038] S2.33: The obtained objective function value is transformed through mapping to obtain the fitness value of each particle. The formula is:

[0039]

[0040] in, represents the fitness function, θ represents the weight factor, ε represents the correction value, and ψ represents the offset.

[0041] Specifically, the specific steps of S2.3 also include:

[0042] S2.34: According to the fitness value of each particle, the position of the particle with the largest fitness value is set as the global optimal solution, and the parent generation is selected based on the global optimal solution to perform the crossover operation of the genetic algorithm. After the crossover operation, a new candidate strategy for the offspring is generated, and the newly generated offspring is randomly mutated;

[0043] S2.35: Update the speed and position of the particle, calculate the fitness value of the new position, and update the optimal solution of each particle and the global optimal solution according to the fitness value of the new position;

[0044] S2.36: Combine all particles processed by steps S2.34 and S2.35 to generate a new candidate strategy set;

[0045] S2.37: Perform Pareto front evaluation on all candidate strategies, identify non-inferior solutions, and form the current multi-objective optimization strategy set from the non-inferior solutions.

[0046] Specifically, an electronic device includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of a boiler system optimization energy saving method based on energy balance when executing the computer program.

[0047] Specifically, a computer-readable storage medium stores computer instructions, which, when executed, execute steps of a method for optimizing energy saving of a boiler system based on energy balance.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] 1. The present invention proposes a boiler system optimization and energy-saving method based on energy balance. The operating data of the boiler system is collected in real time by sensors, and a three-dimensional thermal flow field model is constructed and trained using CFD methods and deep learning technology, which can more accurately predict exhaust heat loss and other key performance indicators. A method combining genetic algorithm and particle swarm optimization algorithm is adopted to generate a multi-objective optimization strategy set, which can comprehensively consider multiple performance indicators, such as thermal efficiency, exhaust heat loss, and fuel consumption, and formulate more comprehensive and optimized energy-saving measures.

[0050] 2. The present invention proposes a boiler system optimization and energy-saving method based on energy balance, establishes a closed-loop feedback mechanism, monitors the execution effect of energy-saving optimization measures in real time, and dynamically adjusts the optimization strategy according to the actual execution effect to ensure that the optimization measures are always consistent with the actual operating conditions of the boiler system, thereby achieving continuous and efficient energy saving; energy balance analysis and data tracking are performed to ensure that the system is always in the best operating state, potential problems are discovered and resolved in a timely manner, and the energy-saving effect is further improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 A schematic diagram of a boiler system energy-saving optimization method based on energy balance according to the present invention;

[0052] Figure 2 The following is a working diagram of the method for optimizing energy saving of a boiler system based on energy balance of the present invention;

[0053] Figure 3 A three-dimensional thermal flow field model prediction flow chart of the boiler system optimization and energy saving method based on energy balance of the present invention;

[0054] Figure 4 This is a multi-objective optimization flow chart of the boiler system energy-saving optimization method based on energy balance of the present invention. DETAILED DESCRIPTION

[0055] Example 1

[0056] See also Figure 1-Figure 3 , an embodiment provided by the present invention: a boiler system optimization energy saving method based on energy balance, comprising the following steps:

[0057] Step S1: The operation data of the boiler system is collected in real time through sensors and preprocessed. Based on the preprocessed operation data, a three-dimensional thermal flow field model of the boiler is constructed using the CFD method, and a real-time data update mechanism is introduced to dynamically simulate the heat distribution during the exhaust process. The three-dimensional thermal flow field model of the boiler is trained by combining the deep learning method and physical rule constraints to obtain the prediction result of exhaust heat loss;

[0058] The operating data of the boiler system include: 1) temperature data, such as furnace temperature, flue gas temperature, and water temperature; 2) pressure data, such as main steam pressure and flue gas pressure; 3) flow data, such as feed water flow, steam flow, and flue gas flow; 4) fuel data, such as coal quantity or fuel oil quantity, and coal quality analysis data; 5) flue gas composition data, such as flue gas oxygen content; 6) grate and burner data, such as grate speed and burner status.

[0059] Step S2: setting a reduction target value for boiler exhaust heat loss, using a genetic algorithm combined with a particle swarm optimization algorithm to generate a multi-objective optimization strategy set, and formulating energy-saving optimization measures based on the multi-objective optimization strategy set;

[0060] Setting the reduction target value for boiler exhaust heat loss includes: referring to industry standards or expert recommendations; setting specific target values ​​based on the actual operating conditions of the boiler and energy conservation and emission reduction requirements. For example, if the current exhaust heat loss is 6%, the target value can be set to reduce it to 4% or lower.

[0061] Step S3: Establish a closed-loop feedback mechanism to monitor the execution effect of energy-saving optimization measures in real time, and dynamically adjust the multi-objective optimization strategy according to the execution effect;

[0062] Furthermore, in step S3, a reinforcement learning method is used to dynamically adjust the multi-objective optimization strategy, and the specific steps include:

[0063] A1: The operating state of the boiler system is defined as the state space, and the control parameters are defined as the action space;

[0064] A2: Design a reward function to evaluate the impact of each control parameter adjustment on system performance;

[0065] A3: Select the Q-learning reinforcement learning algorithm to perform training using real-time data and reward functions, wherein the Q-learning reinforcement learning algorithm is a prior art in the art and is not an inventive solution of the present application, and is not described in detail here;

[0066] A4: Set an update cycle and regularly update the multi-objective optimization strategy based on the output results of the reinforcement learning algorithm.

[0067] Step S4: After the energy-saving optimization measures are implemented, the performance indicators of the boiler system are tested and evaluated, and a real-time feedback mechanism is established. The optimization effect is evaluated in real time according to the monitoring data to determine whether the optimization effect reaches the expected goal. If the expected goal is not reached, a genetic algorithm is used for iterative optimization, and the multi-objective optimization strategy set is updated. The iterative optimization process of the genetic algorithm is the prior art content in this field, and is not the inventive solution of the present application, and will not be described in detail here.

[0068] Step S5: Regularly perform energy balance analysis on the boiler system to identify potential energy-saving points, and establish a data tracking mechanism to continuously monitor the operation status and energy-saving effect of the boiler system, and regularly evaluate the energy-saving effect.

[0069] The specific steps of step S1 include:

[0070] S1.1: Use sensors to collect the operating data of the boiler system in real time and perform preprocessing to obtain the preprocessed operating data X = {x 1 ,...,x n}, where x n Indicates the nth running data, where n indicates the number of running data;

[0071] S1.2: Based on X = {x 1 ,...,x n}, using CFD method to build a three-dimensional model of the boiler according to the geometric structure of the boiler, using momentum conservation equations to obtain fluid flow and energy transfer process, combining the law of conservation of energy to establish the heat balance equation, and X = {x 1 ,…,x n} Input the three-dimensional model as boundary conditions to build a three-dimensional thermal flow field model of the boiler;

[0072] The specific steps of using CFD method to build a three-dimensional model of the boiler include:

[0073] B1: Obtain the actual size, structure, and material data of the boiler, and use CAD or CFD pre-processing software to create a three-dimensional model of the boiler;

[0074] B2: Divide the 3D model into computational units and verify the mesh independence;

[0075] B3: Set boundary conditions based on the preprocessed operating data, including inlet velocity, temperature, and pressure, and select a turbulence model based on the characteristics of fluid flow and heat transfer in the boiler;

[0076] B4: According to the complexity of the problem and the computing resources, set the solution parameters of the CFD solver, such as the number of iterations and convergence criteria;

[0077] B5: Use the CFD solver to solve the turbulence model to obtain the simulation results of fluid flow and heat transfer, and monitor the convergence of the solution and the consumption of computing resources during the calculation process;

[0078] B6: Use CFD post-processing software to visualize the calculation results, such as drawing cloud maps or vector maps of velocity field, temperature field, and pressure field, and analyze and interpret the calculation results to extract information for optimal design or operation control of the boiler.

[0079] S1.3: Introduce a real-time data update mechanism to dynamically simulate the heat distribution in the exhaust process based on real-time operation data. Based on the simulation results, the three-dimensional thermal flow field model of the boiler is trained in combination with a deep learning method. The specific steps of the deep learning method include:

[0080] (1) Obtain historical operation data and simulation result data, build a fusion data set, and perform data preprocessing and feature engineering;

[0081] (2) Select a convolutional neural network model and use the fused data to train the convolutional neural network model;

[0082] (3) Use gradient descent algorithm and back propagation algorithm for optimization;

[0083] (4) Use the validation set or test set to evaluate the trained model and optimize the model based on the evaluation results, such as adjusting the model structure and parameter settings;

[0084] (5) The trained deep learning model is integrated into the three-dimensional thermal flow field simulation system to predict and optimize the heat distribution and flow conditions during the smoke exhaust process.

[0085] S1.4: Set the loss function, add the constraints of energy conservation, heat conduction equation, and momentum conservation to the loss function, and obtain the exhaust heat loss prediction results of the boiler three-dimensional thermal flow field model based on real-time operation data.

[0086] Among them, the conservation of energy, the heat conduction equation, and the momentum conservation formula are respectively the existing technical contents in this field, which are not the creative solutions of this application and will not be elaborated here.

[0087] The heat balance equation in S1.2 is:

[0088]

[0089] Among them, S los represents exhaust heat loss, S fir represents the heat released by fuel combustion, ρ represents the fluid density, that is, the mass of the fluid per unit volume, and u i and u j They all represent the velocity components of the fluid in the x-axis, y-axis, and z-axis directions, and i,j∈[1,3],i≠j,y i represents the spatial coordinates of the fluid in the x-axis, y-axis, and z-axis directions, τ ij represents the component of the viscous stress tensor, represents the component of the heat flux vector caused by heat conduction in the i direction, It represents the component of the heat flux vector caused by radiation in the i direction, E represents the internal energy and kinetic energy of the fluid per unit mass, and reflects the total energy of the microscopic particle movement and macroscopic movement in the fluid, t represents the current time, h represents the pressure, It represents the time rate of change of energy, and (ρE+h) represents the sum of internal energy and pressure potential energy.

[0090] It should be noted that, in the present invention, i and j can take the values ​​of 1, 2, and 3, corresponding to the directions of the x, y, and z axes, respectively, and u i It represents the velocity component of the fluid in the i direction, that is, the velocity component in one of the x, y, and z directions.

[0091] The loss function formula after adding the constraint term in S1.4 is:

[0092]

[0093] Among them, L tol represents the loss function after adding constraints, represents the prediction result of the kth running data, λ represents the weight of the regularization term, L 1 (T,ρ) represents the loss function L 1 Regularization term, T represents temperature, L en represents the energy conservation constraint, γ 1 represents the weight of the energy conservation constraint, L he represents the constraint term of the heat conduction equation, γ 2 represents the weight of the constraint term of the heat conduction equation, L mo represents the momentum conservation constraint, γ 3 Represents the weight of the momentum conservation constraint.

[0094] The purpose of introducing the regularization term in the present invention is to prevent overfitting and enable the model to maintain a certain generalization ability while fitting the data.

[0095] Example 2

[0096] See also Figure 4 In this embodiment, the specific steps of step S2 include:

[0097] S2.1: Set the target value for reducing exhaust heat loss, and obtain X = {x 1 ,...,x n} data, and define the decision variables and objective function, the formula is:

[0098] f(z)=S los (z 1 ,...,z l )

[0099] Where f(z) represents the objective function of minimizing the heat loss of smoke exhaust, (z 1 ,...,z l ) represents the decision variable, z l represents the lth decision variable, S los (·) represents the exhaust heat loss function;

[0100] The decision variables include: control parameters of the combustion process, such as the fuel-air mixing ratio, combustion temperature, and combustion time; operating parameters of the boiler, such as coal feed, air volume, furnace temperature, and the type and quantity of energy-saving technologies or materials used, where each decision variable represents a controllable factor.

[0101] Furthermore, when setting the target value for reducing exhaust heat loss, it is first necessary to determine a set of initial decision variable values ​​based on historical operating data and system characteristics. Then, the operating data of the boiler system is collected in real time through sensors to evaluate the exhaust heat loss of the current system and compare it with the target value. If the current exhaust heat loss is higher than the target value, it is necessary to adjust the value of the decision variable based on the operating data and system characteristics to optimize the objective function and reduce the exhaust heat loss.

[0102] S2.2: Based on the three-dimensional thermal flow field model established in S1.2 and the physical rule constraints in S1.4, construct the constraints of the optimization problem. The constraints include:

[0103]

[0104] Among them, p 1 represents boiler input energy, p 2 represents boiler output energy, β represents thermal conductivity, ▽T represents temperature gradient, T wal represents the wall temperature, T wal,sp Indicates the maximum temperature that the wall can withstand, F fu Indicates the fuel flow rate, F fu,min Indicates the minimum fuel flow rate, F fu,max Indicates the maximum fuel flow rate.

[0105] The specific steps of step S2 also include:

[0106] S2.3: Generate a set of multi-objective optimization strategies using a hybrid strategy of genetic algorithm and particle swarm optimization algorithm, and evaluate the generated multi-objective optimization strategies using the Pareto frontier method;

[0107] S2.4: Determine whether the evaluation result meets the multi-objective optimization conditions. If not, iteratively optimize the multi-objective optimization strategy to obtain the optimal multi-objective optimization strategy;

[0108] S2.5: Based on the optimal multi-objective optimization strategy, formulate energy-saving optimization measures, including adjusting combustion parameters, optimizing boiler structure, and using energy-saving equipment.

[0109] The specific steps of S2.3 include:

[0110] S2.31: Randomly initialize the position and speed of the particle swarm, set the size of the particle swarm, the number of iterations, the crossover and mutation operation parameters of the genetic algorithm, and set the current position of each particle as its individual optimal solution;

[0111] S2.32: For each candidate strategy in the genetic algorithm and the particle swarm optimization algorithm, a three-dimensional thermal flow field model of the boiler is used for simulation, and the heat loss corresponding to each candidate strategy is extracted from the simulation results, and the extracted heat loss data is input into f(z) to obtain the objective function value of each candidate strategy;

[0112] The methods for obtaining each candidate strategy include:

[0113] (1) For the particle swarm optimization algorithm, the position of each particle is a candidate strategy. At initialization, these positions are randomly generated; during the iteration process, they are updated according to the particle's velocity and the individual / global optimal solution;

[0114] (2) For genetic algorithms, each individual is a candidate strategy. During initialization, the genes of these individuals are randomly generated. During the iteration process, they generate new individuals, i.e., new candidate strategies, through crossover and mutation operations.

[0115] S2.33: The obtained objective function value is transformed through mapping to obtain the fitness value of each particle. The formula is:

[0116]

[0117] in, represents the fitness function, θ represents the weight factor, which is used to adjust the weight of heat loss in the fitness value calculation. When θ is larger, the heat loss has a greater impact on the fitness value. When θ is smaller, the heat loss has a smaller impact on the fitness value. ε represents the correction value, which is used to avoid the situation where the divisor is zero and ensure that the fitness value will not be too large when the exhaust heat loss is close to zero. ψ represents the offset, which is used to adjust the overall level of the fitness value. When ψ is larger, the fitness value will decrease overall. When ψ is smaller, the fitness value will increase overall.

[0118] The specific steps of S2.3 also include:

[0119] S2.34: According to the fitness value of each particle, the position of the particle with the largest fitness value is set as the global optimal solution, and the parent generation is selected based on the global optimal solution to perform the crossover operation of the genetic algorithm. After the crossover operation, a new candidate strategy for the offspring is generated, and the newly generated offspring is randomly mutated;

[0120] S2.35: Update the speed and position of the particle, calculate the fitness value of the new position, and update the optimal solution of each particle and the global optimal solution according to the fitness value of the new position;

[0121] S2.36: Combine all particles processed by steps S2.34 and S2.35 to generate a new candidate strategy set;

[0122] S2.37: Perform Pareto front evaluation on all candidate strategies, identify non-inferior solutions, and form the current multi-objective optimization strategy set from the non-inferior solutions.

[0123] Furthermore, the specific steps of S2.37 include:

[0124] (1) Obtain the generated new candidate strategy set, and for each candidate strategy, calculate the value of the candidate strategy on the objective function;

[0125] (2) For any two candidate strategies A and B;

[0126] If A is not worse than B in all objectives and is better than B in at least one objective, then A is said to Pareto dominate B.

[0127] If a candidate strategy is not Pareto dominated by any strategy, it is a non-inferior solution or a Pareto optimal solution;

[0128] (3) By comparing all candidate strategies, all non-inferior solutions are identified and collected to form a multi-objective optimization strategy set.

[0129] Example 3

[0130] An electronic device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of a method for optimizing energy saving of a boiler system based on energy balance are implemented.

[0131] A computer-readable storage medium stores computer instructions, which, when executed, execute steps of a method for optimizing energy saving of a boiler system based on energy balance.

[0132] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in the field may also change, modify, replace and modify the above-mentioned embodiments without departing from the scope of protection of the purpose of the present invention and the claims, and all of these are within the protection of the present invention.

Claims

1. A boiler system optimization and energy saving method based on energy balance, characterized in that: include: Step S1: The operation data of the boiler system is collected in real time through sensors and preprocessed. Based on the preprocessed operation data, a three-dimensional thermal flow field model of the boiler is constructed using the CFD method, and a real-time data update mechanism is introduced to dynamically simulate the heat distribution during the exhaust process. The three-dimensional thermal flow field model of the boiler is trained by combining the deep learning method and physical rule constraints to obtain the prediction result of exhaust heat loss; Step S2: setting a reduction target value for boiler exhaust heat loss, using a genetic algorithm combined with a particle swarm optimization algorithm to generate a multi-objective optimization strategy set, and formulating energy-saving optimization measures based on the multi-objective optimization strategy set; Step S3: Establish a closed-loop feedback mechanism to monitor the execution effect of energy-saving optimization measures in real time, and dynamically adjust the multi-objective optimization strategy according to the execution effect; Step S4: After the energy-saving optimization measures are implemented, the performance indicators of the boiler system are tested and evaluated, and a real-time feedback mechanism is established to evaluate the optimization effect in real time based on the monitoring data to determine whether the optimization effect has achieved the expected goal. If the expected goal has not been achieved, a genetic algorithm is used for iterative optimization, and the multi-objective optimization strategy set is updated; Step S5: regularly conduct energy balance analysis on the boiler system to find potential energy-saving points, establish a data tracking mechanism, continuously monitor the operation status and energy-saving effect of the boiler system, and regularly evaluate the energy-saving effect; The specific steps of step S1 include: S1.1: Use sensors to collect the operating data of the boiler system in real time and perform preprocessing to obtain the preprocessed operating data X = {x1,…,x n }, where x n Indicates the nth running data, where n indicates the number of running data; S1.2: Based on X={x1,...,x n }, using CFD method to build a three-dimensional model of the boiler according to the geometric structure of the boiler, using momentum conservation equations to obtain fluid flow and energy transfer process, combining the law of conservation of energy to establish the heat balance equation, and X = {x1,…,x n } Input the three-dimensional model as boundary conditions to build a three-dimensional thermal flow field model of the boiler; S1.3: Introduce a real-time data update mechanism to dynamically simulate the heat distribution during the exhaust process based on real-time operation data. Based on the simulation results, train the three-dimensional thermal flow field model of the boiler in combination with deep learning methods. S1.4: Set the loss function, add the constraints of energy conservation, heat conduction equation and momentum conservation to the loss function, and obtain the exhaust heat loss prediction result of the boiler three-dimensional thermal flow field model based on real-time operation data; The heat balance equation in S1.2 is: Among them, S los represents exhaust heat loss, S fir represents the heat released by fuel combustion, ρ represents the fluid density, u i and u j They all represent the velocity components of the fluid in the x-axis, y-axis, and z-axis directions, and i,j∈[1,3],i≠j,y i represents the spatial coordinates of the fluid in the x-axis, y-axis, and z-axis directions, τ ij represents the component of the viscous stress tensor, represents the component of the heat flux vector caused by heat conduction in the i direction, It represents the component of the heat flux vector caused by radiation in the i direction, E represents the internal energy and kinetic energy of the fluid per unit mass, t represents the current time, and h represents the pressure.

2. The method for optimizing and energy saving of a boiler system based on energy balance according to claim 1, characterized in that: The loss function formula after adding the constraint term in S1.4 is: Among them, L tol represents the loss function after adding constraints, represents the prediction result of the kth running data, λ represents the weight of the regularization term, L1(T,ρ) represents the L1 regularization term of the loss function, T represents the temperature, L en represents the energy conservation constraint term, γ1 represents the weight of the energy conservation constraint term, L he represents the constraint term of the heat conduction equation, γ2 represents the weight of the constraint term of the heat conduction equation, L mo represents the momentum conservation constraint term, and γ3 represents the weight of the momentum conservation constraint term.

3. The method for optimizing and energy saving of a boiler system based on energy balance according to claim 2, characterized in that: The specific steps of step S2 include: S2.1: Set the target value for reducing exhaust heat loss, and obtain X = {x1,…,x n } data, and define the decision variables and objective function, the formula is: f(z)=S los (z1,...,z l ) Where f(z) represents the objective function of minimizing the heat loss of smoke exhaust, (z1,...,z l ) represents the decision variable, z l represents the lth decision variable, S los (·) represents the exhaust heat loss function; S2.2: Based on the three-dimensional thermal flow field model of the boiler established in S1.2 and the physical rule constraints in S1.4, construct the constraints of the optimization problem. The constraints include: Among them, p1 represents the boiler input energy, p2 represents the boiler output energy, β represents the thermal conductivity, represents the temperature gradient, T wal represents the wall temperature, T wal,sp Indicates the maximum temperature accepted by the wall, F fu Indicates the fuel flow rate, F fu,min Indicates the minimum fuel flow rate, F fu,max Indicates the maximum fuel flow rate.

4. The method for optimizing and energy saving of a boiler system based on energy balance according to claim 3, characterized in that: The specific steps of step S2 also include: S2.3: Generate a set of multi-objective optimization strategies using a hybrid strategy of genetic algorithm and particle swarm optimization algorithm, and evaluate the generated multi-objective optimization strategies using the Pareto frontier method; S2.4: Determine whether the evaluation result meets the multi-objective optimization conditions. If not, iteratively optimize the multi-objective optimization strategy to obtain the optimal multi-objective optimization strategy; S2.5: Develop energy-saving optimization measures based on the optimal multi-objective optimization strategy.

5. The method for optimizing and energy saving of a boiler system based on energy balance according to claim 4, characterized in that: The specific steps of S2.3 include: S2.31: Randomly initialize the position and speed of the particle swarm, set the size of the particle swarm, the number of iterations, the crossover and mutation operation parameters of the genetic algorithm, and set the current position of each particle as its individual optimal solution; S2.32: For each candidate strategy in the genetic algorithm and the particle swarm optimization algorithm, a three-dimensional thermal flow field model of the boiler is used for simulation, and the heat loss corresponding to each candidate strategy is extracted from the simulation results, and the extracted heat loss data is input into f(z) to obtain the objective function value of each candidate strategy; S2.33: The obtained objective function value is transformed through mapping to obtain the fitness value of each particle. The formula is: in, represents the fitness function, θ represents the weight factor, ε represents the correction value, and ψ represents the offset.

6. The method for optimizing and energy saving of a boiler system based on energy balance according to claim 5, characterized in that: The specific steps of S2.3 also include: S2.34: According to the fitness value of each particle, the position of the particle with the largest fitness value is set as the global optimal solution, and the parent generation is selected based on the global optimal solution to perform the crossover operation of the genetic algorithm. After the crossover operation, a new candidate strategy for the offspring is generated, and the newly generated offspring is randomly mutated; S2.35: Update the speed and position of the particle, calculate the fitness value of the new position, and update the optimal solution of each particle and the global optimal solution according to the fitness value of the new position; S2.36: Combine all particles processed by steps S2.34 and S2.35 to generate a new candidate strategy set; S2.37: Perform Pareto frontier evaluation on all candidate strategies, identify non-inferior solutions, and form the current multi-objective optimization strategy set from the non-inferior solutions.

7. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the boiler system optimization and energy saving method based on energy balance described in any one of claims 1 to 6 are implemented.

8. A computer-readable storage medium, characterized in that: Computer instructions are stored thereon, and when the computer instructions are executed, the steps of the method for optimizing energy saving of a boiler system based on energy balance as described in any one of claims 1 to 6 are executed.

Images