An information adaptive fusion industrial process energy consumption optimization method

Abstract

The embodiment of the present disclosure provides an information adaptive fusion industrial process energy consumption optimization method, which belongs to the technical field of computing, and specifically comprises the following steps: step 1, obtaining key process parameters related to energy consumption and data related to energy consumption cost in an industrial process to form industrial process characteristics; step 2, taking minimum energy consumption and minimum energy consumption cost as targets, constructing a constraint condition with the industrial process characteristics, and establishing an industrial process energy consumption optimization model; and step 3, solving the industrial process energy consumption optimization model by using an information adaptive fusion multi-objective evolutionary algorithm to obtain a plurality of groups of optimal key process parameters with the lowest energy consumption and energy consumption cost. Through the scheme of the present disclosure, the optimization efficiency and environmental protection are improved.

Description

Technical Field

[0001] This disclosure relates to the field of computing technology, and in particular to an industrial process energy consumption optimization method based on adaptive information fusion. Background Technology

[0002] Currently, process industries transform minerals into high-purity products through complex physicochemical reactions. These material conversion processes involve energy-intensive steps such as electrolysis, making them typical high-energy-consuming industries and reflecting their heavy reliance on energy. Therefore, the need for energy conservation and efficiency improvement in process industries is urgent. Against the backdrop of the accelerated intelligent and green transformation of process industries, optimizing energy consumption in typical high-energy-consuming production processes is beneficial for releasing the emission reduction potential of process industries, promoting their green and low-carbon transformation, and is also an important measure to support the high-quality development of process industries.

[0003] With the development of artificial intelligence technology, research on energy consumption optimization problems in industrial processes, both domestically and internationally, can be divided into traditional optimization methods and heuristic intelligent optimization methods. Traditional optimization methods are optimization techniques based on classical mathematical theories and algorithms, relying on deterministic algorithms and existing mathematical theories. They are suitable for optimization problems with specific structures and mainly include integer programming methods and duality theory. Traditional optimization methods perform well in dealing with linear optimization and small-scale optimization problems. However, industrial process optimization problems generally exhibit highly nonlinear, complex constraints, and multi-objective characteristics; these problems may even lack explicit analytical solutions. Against this backdrop, traditional optimization methods struggle to meet the stringent requirements of industrial optimization control in terms of solution efficiency and robustness. Heuristic intelligent optimization methods draw on algorithms developed in nature and biological evolution, emphasizing the search for global optimal solutions through adaptive mechanisms and swarm intelligence. They are suitable for complex, nonlinear, and multimodal problems, possessing strong flexibility and adaptability. These mainly include particle swarm optimization and evolutionary algorithms, and are currently the mainstream methods. However, most current research still has the following shortcomings: 1) The global optimization capability of the algorithm is difficult to meet the requirements of green, low-carbon and high-quality development of industrial processes, and needs to be further improved; 2) The unconstrained Pareto front and the constrained Pareto front respectively contain rich optimization objectives and constraint information, but they have not been fully explored, utilized and integrated, making it difficult to balance the feasibility and convergence of the solution in the optimization process.

[0004] It is evident that there is an urgent need for an industrial process energy consumption optimization method that features highly efficient and environmentally friendly adaptive information fusion. Summary of the Invention

[0005] In view of this, the present disclosure provides an industrial process energy consumption optimization method based on adaptive information fusion, which at least partially solves the problems of insufficient global optimization capability and inadequate utilization of optimization target and constraint information in the prior art.

[0006] This disclosure provides an industrial process energy consumption optimization method based on adaptive information fusion, including:

[0007] Step 1: Obtain key process parameters related to energy consumption and data related to energy consumption costs in the industrial process to form industrial process characteristics;

[0008] Step 2: With minimum energy consumption and minimum energy cost as objectives, construct constraints based on the characteristics of the industrial process, and establish an industrial process energy consumption optimization model;

[0009] Step 3: Use the multi-objective evolutionary algorithm with information adaptive fusion to solve the industrial process energy consumption optimization model and obtain multiple sets of optimal key process parameters with the lowest energy consumption and energy cost.

[0010] According to a specific implementation of an embodiment of this disclosure, step 2 specifically includes:

[0011] Step 2.1: Establish the first objective function with the goal of minimizing power consumption.

[0012]

[0013] in, This indicates the number of electrolytic cells in the first / second cathode cycle during copper electrolysis. Indicates the number of time-of-use electricity price ranges. Indicates the first Duration of each time-of-use electricity price range Indicates the first Current in each time-of-use pricing zone This indicates the first cathode cycle electrolytic cell, the... The voltage of the slot in each time-of-use electricity price zone This indicates the second cathode cycle electrolytic cell, the first... The voltage of the slot in each time-of-use electricity price zone;

[0014] Step 2.2: Establish a second objective function with the goal of minimizing electricity consumption cost.

[0015]

[0016] in, Indicates the first Electricity unit price within each time-of-use electricity price range;

[0017] Step 2.3: Establish daily output constraints based on the characteristics of the industrial process.

[0018] ;

[0019] Step 2.4: Establish current regulation constraints based on industrial process characteristics.

[0020] ;

[0021] Step 2.5: Establish copper-acid relationship constraints based on industrial process characteristics.

[0022]

[0023] in, Indicates the first Copper ion concentration in each time-of-use electricity price range Indicates the first The concentration of acid radicals in each time-of-use electricity price range;

[0024] Step 2.6: Input the sample set into the trained slot voltage prediction model to obtain the calculation relationship between slot voltage and each parameter.

[0025]

[0026] in, Indicates the first Electrolyte temperature for each time-of-use electricity price range This indicates the cathode period in which the electrolysis system is located. Indicates the concentration of impurity metal ions;

[0027] Step 2.7: Establish boundary constraints corresponding to copper ion concentration, acid anion concentration, current, and electrolyte temperature based on the characteristics of the industrial process.

[0028] ;

[0029] Step 2.8: Establish an industrial process energy consumption optimization model based on the first objective function, the second objective function, daily output constraints, current regulation constraints, copper-acid relationship constraints, the calculated relationships between cell voltage and various parameters, and boundary constraints.

[0030] .

[0031] According to a specific implementation of an embodiment of this disclosure, step 3 specifically includes:

[0032] Step 3.1: Randomly initialize the population to form a shared population. The parent population of each individual ;

[0033] Step 3.2: Based on the simulated binary crossover operator and the polynomial mutation operator, generate a shared... The offspring population of each individual Merging parent populations and offspring population to form a new population ;

[0034] Step 3.3, based on the constrained dominance principle, the new population is... Divided into A different Pareto frontier And calculate the number of individuals in different Pareto fronts;

[0035] Step 3.4, Calculate the new population Feasibility Then determine whether the current stage is a feasible stage. If not, proceed to step 3.5; if so, proceed to step 3.6.

[0036] Step 3.5: If the current stage is infeasible, use a clustering-based individual selection strategy to select from a new population. Screening Individuals form a new parental population. ;

[0037] Step 3.6: If the current stage is a feasible stage, calculate the new population using the fitness function of information adaptive fusion. The fitness of each individual is assessed, and the individual with the lowest fitness is selected. Individuals form a new parental population. ;

[0038] Step 3.7: Repeat steps 3.2 to 3.6 until the termination condition is met, and obtain multiple sets of optimal key process parameters with the lowest energy consumption and energy cost.

[0039] According to a specific implementation of an embodiment of this disclosure, the step of determining whether the current stage is a feasible stage includes:

[0040] Determine if the feasibility rate is greater than 0;

[0041] If the feasibility rate is greater than 0, then the current stage is determined to be a feasible stage;

[0042] If the feasibility rate is less than 0, the current stage is determined to be an infeasible stage.

[0043] According to a specific implementation of an embodiment of this disclosure, step 3.5 specifically includes:

[0044] Step 3.5.1, find the first A cutting-edge, making The sum of the frontier individuals is less than , recorded as , The sum of the frontier individuals is greater than and will All leading individuals were placed into the new parent population. ;

[0045] Step 3.5.2, based on the K-means clustering algorithm, the first... The frontier individuals are divided into There are 10 categories, and the cluster center is denoted as _____. ;

[0046] Step 3.5.3, set the ideal point Connected to the cluster centers to form vectors Calculate the first frontier individual and vector The distance is calculated using the nearest distance principle. The corresponding number of individuals at the first frontier ;

[0047] Step 3.5.4: Sort the clusters from smallest to largest based on the number of individuals in the first frontier, and record the corresponding cluster centers and cluster numbers as follows: and ;

[0048] Step 3.5.5, compare sequentially and The size relationship, if Less than Then All corresponding numbers Individuals at the forefront are introduced into the new parent population. ,and Otherwise from All corresponding numbers Random selection from frontier individuals Each individual is placed into a new parent population. .

[0049] According to a specific implementation of an embodiment of this disclosure, step 3.6 specifically includes:

[0050] Step 3.6.1: Based on the constrained dominance principle, calculate the new population while considering the constraints. The ranking value of each individual To obtain constrained Pareto front information;

[0051] Step 3.6.2, Calculation based on congestion distance Crowding of each individual in the new population ;

[0052] Step 3.6.3: Based on the Pareto dominance principle, calculate the new population without considering any constraints. The ranking value of each individual To obtain information on the unconstrained Pareto front;

[0053] Step 3.6.4: Adaptively fuse constrained Pareto front information and unconstrained Pareto front information based on feasibility rate, and calculate the population based on the fitness function. Fitness value of each individual

[0054]

[0055] in, This represents the maximum crowding level for all individuals.

[0056] Step 3.6.5: Select the least fit in sequence. Individuals form a new parental population. .

[0057] The information adaptive fusion industrial process energy consumption optimization scheme in this embodiment includes: Step 1, acquiring key process parameters related to energy consumption and data related to energy consumption cost in the industrial process to form industrial process characteristics; Step 2, establishing an industrial process energy consumption optimization model with minimum energy consumption and minimum energy cost as objectives and constructing constraints based on the industrial process characteristics; Step 3, solving the industrial process energy consumption optimization model using a multi-objective evolutionary algorithm of information adaptive fusion to obtain multiple sets of optimal key process parameters with the lowest energy consumption and energy cost.

[0058] The beneficial effects of this disclosure are as follows: The scheme of this disclosure determines multiple key process parameters related to energy consumption and data related to energy cost in industrial processes. Then, it analyzes the characteristics of the industrial process, establishes an industrial process energy consumption optimization model with the objectives of minimum energy consumption and minimum energy cost, and uses the characteristics of the industrial process as constraints. Finally, it uses an information-adaptive fusion multi-objective evolutionary algorithm to solve the industrial process energy consumption optimization model, obtaining the optimal key process parameters with the lowest energy consumption and energy cost. Specifically, in the process of solving the industrial process energy consumption optimization model using the information-adaptive fusion multi-objective evolutionary algorithm, the Pareto dominance principle (DP) and the constrained dominance principle (CDP) are used to mine the optimization objectives and constraints implicit in the unconstrained and constrained Pareto fronts, respectively. The feasibility rate is used as the population evolution state to adaptively fuse the two information, constructing an adaptive fitness function, achieving a balance between the feasibility and convergence of the solution during the optimization process. Simultaneously, the crowding information in the fitness function increases the diversity of solutions. Furthermore, in the infeasibility phase, the relationship between the first frontier individuals and the Lth frontier is analyzed by clustering algorithm, and the Lth frontier individuals are selected from the perspective of diversity. This balances the diversity and feasibility of the solution in this phase, and improves the optimization efficiency and environmental friendliness. Attached Figure Description

[0059] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0060] Figure 1 A flowchart illustrating an industrial process energy consumption optimization method based on adaptive information fusion, provided in an embodiment of this disclosure;

[0061] Figure 2 This is a schematic diagram of a copper electrolysis production process provided in an embodiment of the present disclosure;

[0062] Figure 3 This is a schematic diagram of a multi-objective evolutionary algorithm for adaptive information fusion provided in an embodiment of the present disclosure. Detailed Implementation

[0063] The embodiments of this disclosure will now be described in detail with reference to the accompanying drawings.

[0064] The following specific examples illustrate the implementation of this disclosure. Those skilled in the art can easily understand other advantages and effects of this disclosure from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them. This disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this disclosure. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0065] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this disclosure, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0066] It should also be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of this disclosure. The illustrations only show the components related to this disclosure and are not drawn according to the number, shape and size of the components in actual implementation. In actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0067] Furthermore, specific details are provided in the following description to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that the described aspects can be practiced without these specific details.

[0068] This disclosure provides an industrial process energy consumption optimization method based on adaptive information fusion, which can be applied to the industrial process energy consumption optimization process in industrial production scenarios.

[0069] See Figure 1 This is a schematic flowchart illustrating an industrial process energy consumption optimization method based on adaptive information fusion, provided in an embodiment of this disclosure. Figure 1 As shown, the method mainly includes the following steps:

[0070] Step 1: Obtain key process parameters related to energy consumption and data related to energy consumption costs in the industrial process to form industrial process characteristics;

[0071] In practical implementation, the copper electrolysis production process will be used as an example for illustration. A schematic diagram of the production process is shown below. Figure 2 As shown. Based on the actual production process, it can be seen that the energy involved in the copper electrolysis process includes electricity and steam. The electricity consumption accounts for more than 90%, and the steam consumption accounts for less than 10%. Therefore, this application only considers the electricity consumption.

[0072] Furthermore, according to Joule's law, energy consumption is related to current, voltage, and time. Analysis of the actual copper electrolysis production process and its location reveals that time-of-use electricity pricing in the location of the copper electrolysis enterprise affects the energy usage pattern of the production process, thus influencing energy consumption. Table 1 shows the time-of-use electricity pricing in the location of the copper electrolysis enterprise from July to September. In addition, based on electrochemical principles, the key process parameters affecting the cell voltage in the copper electrolysis process include electrode spacing, current density, electrolyte temperature, copper ion concentration, acid anion concentration, and impurity content.

[0073] Table 1

[0074]

[0075] Step 2: With minimum energy consumption and minimum energy cost as objectives, construct constraints based on the characteristics of the industrial process, and establish an industrial process energy consumption optimization model;

[0076] In practical implementation, to more accurately establish an industrial process energy consumption optimization model, the following model assumptions are made: 1. The current of the electrolytic cell can reach any given value as required; 2. Through the purification process, the concentrations of copper ions and acid radical ions in the electrolytic cell can reach any given value as required and remain stable; 3. The temperature of the electrolytic cell can stably reach any given value as required; 4. Through the purification process, the impurity ions in the electrolytic cell are all in a stable state.

[0077] Furthermore, the energy consumption optimization process for copper electrolysis needs to reduce energy consumption and energy costs under production process constraints such as output, process parameter range, and current adjustment range. This involves establishing a copper electrolysis energy consumption optimization model, specifically including:

[0078] The meanings of the variables in the energy consumption optimization model for the copper electrolysis production process are shown in Table 2.

[0079] Table 2

[0080]

[0081] With minimizing energy consumption as the optimization objective, the objective function expression is established as follows:

[0082]

[0083] Furthermore, with the goal of minimizing electricity consumption cost, the objective function expression is established as follows:

[0084]

[0085] Furthermore, the copper electrolysis production process needs to ensure that the output reaches the target, so a daily output constraint expression is established:

[0086]

[0087] Furthermore, to ensure production stability, the current regulation needs to be controlled within a reasonable range. Therefore, a current regulation constraint expression is established:

[0088]

[0089] Furthermore, to control the surface quality of copper electrolytic products, the relationship between copper ion concentration and acid anion concentration needs to be reasonably controlled. A constraint expression for the copper-acid relationship is established as follows:

[0090]

[0091] Furthermore, 2800 data points from the copper electrolysis production process were collected and divided into training and testing sets at a 4:1 ratio. Under a constant impurity ion concentration, a cell voltage prediction model was established using a gated recurrent neural network (GRU). The loss function was the root mean square error (RMSE) between the predicted and actual cell voltage values. The learning rate was dynamically adjusted using exponential decay, and the root mean square propagation gradient descent method was used as the optimizer. The cell voltage prediction model was trained on the training set, resulting in an RMSE of 0.28 and an R² of 0.92. The trained cell voltage prediction model was tested using the testing set, achieving an RMSE of 0.40. The calculation relationship between cell voltage and various parameters was obtained as follows:

[0092]

[0093] Furthermore, copper ion concentration, anion ion concentration, current, and electrolyte temperature are key decision variables and must satisfy the following boundary constraints:

[0094]

[0095] Furthermore, considering the aforementioned minimum energy consumption and minimum energy cost objectives, as well as the constraints of the copper electrolysis production process, an energy consumption optimization model for the copper electrolysis production process is established:

[0096] .

[0097] Step 3: Use the multi-objective evolutionary algorithm with information adaptive fusion to solve the industrial process energy consumption optimization model and obtain multiple sets of optimal key process parameters with the lowest energy consumption and energy cost.

[0098] In practice, the industrial process energy consumption optimization model is solved using a multi-objective evolutionary algorithm with adaptive information fusion. The algorithm flowchart is as follows: Figure 3 As shown, multiple sets of optimal key process parameters with the lowest energy consumption and energy cost are obtained.

[0099] Based on the above embodiments, step 3 specifically includes:

[0100] Step 31: Randomly initialize the population to form a shared population. The parent population of each individual In specific implementation, Set to 100;

[0101] Step 32: Based on the simulated binary crossover operator and the polynomial mutation operator, generate a shared... The offspring population of each individual Merge the parent populations and the offspring population to form a new population ;

[0102] Step 33: Based on the constrained dominance principle (CDP), the new population is... Divided into A different Pareto frontier And calculate the number of individuals in the different Pareto fronts;

[0103] Step 34, calculate the new population. Feasibility And determine whether the current stage is a feasible stage or an infeasible stage;

[0104] Step 35, if the current stage is an infeasible stage ( A clustering-based individual selection strategy is used to select individuals from the new population. Screening Individuals form a new parental population. ;

[0105] Step 36, if the current stage is a feasible stage ( The fitness function of adaptive information fusion is used to calculate the new population. The fitness of each individual is assessed, and the individual with the lowest fitness is selected. Individuals form a new parental population. ;

[0106] Step 37: Repeat steps 32-36 until the termination condition is met, and obtain multiple sets of optimal key process parameters with the lowest energy consumption and energy cost. In specific implementation, the termination condition is set to the number of function evaluations reaching 15,000.

[0107] Furthermore, step 35 specifically includes:

[0108] Step 351, find the first A cutting-edge, making The sum of the frontier individuals is less than (recorded as) ), The sum of the frontier individuals is greater than and the All leading individuals were placed into the new parent population. ;

[0109] Step 352, based on the K-means clustering algorithm, the first... The frontier individuals are divided into There are 10 categories, and the cluster center is denoted as _____. ;

[0110] Step 353, set the ideal point Connected to the cluster centers to form a vector Calculate the first frontier individual and the vector. The distance is calculated using the nearest distance principle. The corresponding number of the first frontier individuals ;

[0111] Step 354: Sort the clusters from smallest to largest according to the quantities, and record the corresponding cluster centers and cluster counts as follows: and ;

[0112] Step 355, compare sequentially and The size relationship, if Less than Then the above All corresponding numbers Frontier individuals are placed into the new parent population. ,and Otherwise from the above... All corresponding numbers Random selection from frontier individuals Each individual is placed into the new parent population. .

[0113] Furthermore, step 36 specifically includes:

[0114] Step 361: Based on the CDP, calculate the population while considering the constraints. The ranking value of each individual To obtain constrained Pareto front information;

[0115] Step 362, calculate the population based on crowding distance. Crowding level of each individual ;

[0116] Step 363: Based on the Pareto dominance principle (DP), calculate the population without considering constraints. The ranking value of each individual To obtain information on the unconstrained Pareto front;

[0117] Step 364: Adaptively fuse the constrained Pareto front information and the unconstrained Pareto front information based on the feasibility rate, and calculate the population according to the following fitness function. Fitness value of each individual:

[0118]

[0119] in, This represents the maximum crowding level for all individuals.

[0120] In practical implementation, based on the multiple sets of optimal key process parameters, a set of optimal key process parameters with the lowest energy consumption cost is selected. Then, the third-generation non-dominated sorting genetic algorithm (NSGAIII), the conflict-information-based genetic algorithm (NSGAII-conflict), the strong-dominance-based genetic algorithm (NSGAII-SDR), the goal-decomposition-based genetic algorithm (RPD-NSGAII), and the r-dominance-based genetic algorithm (r-NSGAII) are used as comparison algorithms. Each algorithm is run independently 25 times, and the mean and standard deviation are calculated. The experimental results are shown in Table 3, and Table 4 shows a set of optimal key process parameters obtained by the proposed method in 25 experiments. In Table 3, the experimental results of each comparison method are presented in the form of "mean (standard deviation)," where "+" indicates that the mean of the corresponding method is better than the proposed method, "≈" indicates that the mean of the corresponding method is consistent with the proposed method, and "-" indicates that the mean of the corresponding method is worse than the proposed method. "" indicates that the corresponding method failed to find a feasible solution in 25 calculations. As shown in Table 3, NSGAII-SDR, RPD-NSGAII, and r-NSGAII failed to find a feasible solution in 25 experiments. NSGAIII and NSGAII-conflict ranked lower than the proposed method in terms of energy consumption cost and energy consumption index, proving that the global optimization capability of the proposed method is the best among the six methods. By adaptively fusing the optimization objective and constraint information implied by the unconstrained Pareto front and the constrained Pareto front, the feasibility and convergence of the solution in the optimization process are balanced, thus improving the global optimization capability of the proposed method.

[0121] Table 3

[0122]

[0123] Table 4

[0124]

[0125] This embodiment provides an information adaptive fusion-based industrial process energy consumption optimization method. It identifies multiple key process parameters related to energy consumption and data related to energy cost in the industrial process, then analyzes the characteristics of the industrial process. With minimum energy consumption and minimum energy cost as objectives and the industrial process characteristics as constraints, an industrial process energy consumption optimization model is established. Finally, an information adaptive fusion-based multi-objective evolutionary algorithm is used to solve the industrial process energy consumption optimization model, obtaining the optimal key process parameters with the lowest energy consumption and energy cost. Specifically, in solving the industrial process energy consumption optimization model using the information adaptive fusion-based multi-objective evolutionary algorithm, the Pareto dominance principle (DP) and constrained dominance principle (CDP) are used to mine the optimization objectives and constraints implicit in the unconstrained and constrained Pareto fronts, respectively. The feasibility rate is used as the population evolution state to adaptively fuse the two information sets, constructing an adaptive fitness function that achieves a balance between the feasibility and convergence of solutions during the optimization process. Simultaneously, crowding information is incorporated into the fitness function to increase the diversity of solutions. Furthermore, in the infeasibility stage, the relationship between the first frontier individuals and the Lth frontier is analyzed by clustering algorithm, and the Lth frontier individuals are selected from the perspective of diversity. This balances the diversity and feasibility of the solution in this stage, makes full use of the optimization objective and constraint information, and improves the optimization efficiency and environmental friendliness.

[0126] It should be understood that the various parts of this disclosure can be implemented in hardware, software, firmware, or a combination thereof.

[0127] The above description is merely a specific embodiment of this disclosure, but the scope of protection of this disclosure is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this disclosure should be included within the scope of protection of this disclosure. Therefore, the scope of protection of this disclosure should be determined by the scope of the claims.

Claims

1. An industrial process energy consumption optimization method based on adaptive information fusion, characterized in that, include: S1, acquire key process parameters related to energy consumption and data related to energy consumption costs in the industrial process to form industrial process characteristics; S2, with the goals of minimum energy consumption and minimum energy cost, establishes an industrial process energy consumption optimization model by constructing constraints based on the characteristics of the industrial process; S3 utilizes a multi-objective evolutionary algorithm with adaptive information fusion to solve the energy consumption optimization model of industrial processes, obtaining multiple sets of optimal key process parameters with the lowest energy consumption and energy cost; S3 includes: S31, Randomly initialize the population to form a shared... The parent population of each individual ; S32, based on the simulated binary crossover operator and the polynomial mutation operator, generates a shared... The offspring population of each individual Merging parent populations and offspring population to form a new population ; S33, based on the principle of constrained dominance, will create a new population. Divided into A different Pareto frontier Calculate the number of individuals in different Pareto fronts; S34, Calculate the new population feasibility Determine whether the current stage is a feasible stage. If not, execute S35; if so, execute S36. S35, If the current stage is infeasible, a cluster-based individual selection strategy is used to select from the new population. Screening Individuals form a new parental population. ; S36. If the current stage is a feasible stage, use the fitness function of information adaptive fusion to calculate the new population. The fitness of each individual is considered, and the individual with the lowest fitness is selected. Individuals form a new parental population. ; S37, repeat S32 to S36 until the termination condition is met, to obtain multiple sets of optimal key process parameters with the lowest energy consumption and energy cost; S35 includes: Searching for the first A cutting-edge, making The sum of the frontier individuals is less than , recorded as , The sum of the frontier individuals is greater than ,Will All leading individuals were placed into the new parent population. ; Based on the K-means clustering algorithm, the th... The frontier individuals are divided into There are 10 categories, and the cluster center is denoted as _____. ; Ideal point Connected to the cluster centers to form vectors Calculate the first frontier individual and vector The distance is calculated using the nearest distance principle. The corresponding number of individuals at the first frontier ; Sort the clusters by the number of individuals in the first frontier from smallest to largest, and denote the corresponding cluster centers and the number of clusters as follows: and ; Compare in turn and The size relationship, if Less than Then All corresponding numbers Individuals at the forefront are introduced into the new parent population. ,and Otherwise from All corresponding numbers Random selection from frontier individuals Each individual is placed into a new parent population. ; S36 includes: Based on the principle of constrained domination, the new population is calculated considering the constraints. The ranking value of each individual To obtain constrained Pareto front information; Based on crowding distance calculation Crowding of each individual in the new population ; Calculate the new population based on the Pareto dominance principle without considering constraints. The ranking value of each individual To obtain information on the unconstrained Pareto front; The population is calculated based on the fitness function by adaptively fusing constrained and unconstrained Pareto front information according to the feasibility rate. Fitness value of each individual in, This represents the maximum crowding level for all individuals. Select the least fit in sequence Individuals form a new parental population. .

2. The method according to claim 1, characterized in that, S2 specifically includes: S21, with the minimum power consumption as the optimization objective, establish the first objective function. in, This indicates the number of electrolytic cells in the first / second cathode cycle during copper electrolysis. Indicates the number of time-of-use electricity price ranges. Indicates the first Duration of each time-of-use electricity price range Indicates the first Current in each time-of-use pricing zone This indicates the first cathode cycle electrolytic cell, the... The voltage of the slot in each time-of-use electricity price zone This indicates the second cathode cycle electrolytic cell, the first... The voltage of the slot in each time-of-use electricity price zone; S22, with the goal of minimizing electricity consumption cost, establishes a second objective function. in, Indicates the first Electricity unit price for each time-of-use electricity price range; S23, Establish daily output constraints based on industrial process characteristics. ； S24, Establish current regulation constraints based on industrial process characteristics. ； S25, Establishing copper-acid relationship constraints based on industrial process characteristics. in, Indicates the first Copper ion concentration in each time-of-use electricity price range Indicates the first The concentration of acid radicals in each time-of-use electricity price range; S26. Input the sample set into the trained slot voltage prediction model to obtain the calculation relationship between slot voltage and each parameter. in, Indicates the first Electrolyte temperature for each time-of-use electricity price range This indicates the cathode period in which the electrolysis system is located. Indicates the concentration of impurity metal ions; S27, Establish boundary constraints corresponding to copper ion concentration, acid anion concentration, current, and electrolyte temperature based on the characteristics of the industrial process. ； S28. Based on the first objective function, the second objective function, daily output constraints, current regulation constraints, copper-acid relationship constraints, the calculated relationship between cell voltage and various parameters, and boundary constraints, an industrial process energy consumption optimization model is established. 。 3. The method according to claim 1, characterized in that, The step of determining whether the current stage is a feasible stage includes: Determine if the feasibility rate is greater than 0; If the feasibility rate is greater than 0, then the current stage is determined to be a feasible stage; If the feasibility rate is less than 0, the current stage is determined to be an infeasible stage.

Images