A method for optimizing low-carbon manufacturing process parameters of a tubular heat exchanger

Abstract

The application discloses a kind of optimization methods of tubular heat exchanger low-carbon manufacturing process parameters, it is related to tubular heat exchanger manufacturing technical field, including the following steps: according to process, tubular heat exchanger manufacturing process is split, is divided into multiple with carbon emission related process object, the corresponding relationship between process parameters and carbon emission is established, and the before-and-after relationship and parallel relationship between each process are described using AND / OR logic, form process association relationship.The application integrates each process carbon emission calculation by establishing multilevel mapping relationship, realizes whole-process low-carbon collaborative optimization;Combining Sobol method identifies key process parameters, improves parameter screening efficiency and rationality.At the same time, using double population and fuzzy logic regulation mechanism improves optimization ability under complex constraint, the obtained parameter can be directly used for production, realizes the comprehensive optimization of carbon emission, processing time and cost without equipment modification.

Description

Technical Field

[0001] The present invention relates to the technical field of tubular heat exchanger manufacturing, and in particular to a method for optimizing low-carbon manufacturing process parameters of tubular heat exchangers. Background Art

[0002] With the continuous improvement of global climate governance requirements, carbon emission control in the manufacturing industry has become an important direction for promoting green industrial transformation. As a key basic equipment in chemical, energy and other fields, the manufacturing process of tubular heat exchangers usually includes multiple working procedures such as cutting, milling, rounding and welding. The above procedures generally have the characteristics of high energy consumption, complex process parameters and diverse carbon emission sources. Therefore, on the premise of ensuring processing quality and production efficiency, how to reduce carbon emissions in the manufacturing process has become an important technical problem that needs to be solved urgently in the field of tubular heat exchanger manufacturing. At present, research on low-carbon manufacturing mainly focuses on material substitution, energy structure optimization and adjustment of single-process process parameters. For example, cutting parameters are optimized to reduce energy consumption of a single machine, or low-emission welding materials are used to reduce carbon emissions.

[0003] Most of the prior art only optimizes a single process or local link, and lacks systematic analysis of the coupling relationship between multiple processes, which easily leads to the "carbon leakage" phenomenon of local optimization in the overall manufacturing process, that is, the carbon emission reduction effect achieved by one process is offset by the increase in energy consumption of subsequent processes, making it difficult to achieve global optimal carbon emission. Meanwhile, in the actual manufacturing environment, the process is usually affected by multiple strong constraints such as equipment power limitation, processing quality requirements and cost constraints, which makes the problem highly nonlinear and complex. Traditional multi-objective optimization algorithms (such as NSGA-II, particle swarm optimization, etc.) generally have problems such as difficult convergence, low proportion of feasible solutions and deviation between optimization results and actual working conditions when dealing with such strongly constrained problems. In addition, existing methods generally lack explicit modeling of the internal mechanism between process parameters and carbon emissions, and it is difficult to simultaneously meet the multi-objective optimization requirements of carbon emission, processing time and cost from a global perspective. Therefore, there is an urgent need for a low-carbon manufacturing process parameter decision-making method that can decouple the impact of multi-process carbon emissions, quantitatively identify key process parameters, and achieve global optimization under actual constraints.

[0004] The above information disclosed in the Background section is only for enhancing the understanding of the background of the present disclosure, and therefore it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0005] The object of the present invention is to provide a method for optimizing low-carbon manufacturing process parameters of tubular heat exchangers, so as to solve the problems in the above background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for optimizing low-carbon manufacturing process parameters of tubular heat exchangers, comprising the following steps: The manufacturing process of tubular heat exchangers is broken down into multiple carbon emission-related process objects, and the correspondence between process parameters and carbon emissions is established. The AND / OR logic is used to describe the sequential and parallel relationships between each process, forming process association relationships. Based on the process relationships, for each process object, according to the energy consumption and material consumption, a calculation relationship is established with process parameters as input and carbon emissions as output, and the parameters are corrected in combination with production data; Based on the calculation relationship, the Sobol variance decomposition method is used to analyze the impact of each process parameter on the overall carbon emission and overall processing time, identify the main influencing parameters and determine their corresponding weights. Based on the main influencing parameters, a calculation model is established with overall carbon emissions and overall processing time as objectives, and the equipment capacity range, process allowable range and cost range are added as constraints. For the computational model, an optimization method involving two populations is introduced. One type is used to satisfy the constraints, and the other type is used to optimize the objective. The penalty coefficient is dynamically adjusted through fuzzy logic, and individuals are exchanged between populations to obtain multiple sets of candidate parameter combinations. Based on the candidate parameter combinations, target process parameters are selected in conjunction with production requirements and applied to the manufacturing process of tubular heat exchangers.

[0007] The steps to refine the correlation between manufacturing processes and carbon emissions, and improve the accuracy and calculability of expressing multi-process relationships, are as follows: Each process is numbered according to the actual processing sequence, and the dependencies and parallel processing relationships between each process are clearly defined. Based on product quality requirements and manufacturing processes, the process parameters corresponding to different processes are classified and collected. The process parameters of each process are linked and recorded with the corresponding energy consumption and material consumption; By integrating process sequence, parameter information, and carbon emission relationships through a unified expression method, a relational expression result that can be used for calculation is formed.

[0008] The following steps refine the calculation relationship between process parameters and carbon emissions to improve the accuracy of carbon emission calculations: The corresponding energy consumption form is determined based on the operating characteristics of each process, and the energy consumption generated during the processing is calculated. By combining energy carbon emission coefficients, energy consumption is converted into corresponding carbon emissions; The material losses and auxiliary material consumption generated during the processing are statistically analyzed, and the carbon emission coefficient of the materials is introduced for conversion. The carbon emissions from energy consumption and carbon emissions from material consumption are superimposed to obtain the carbon emission results for each process.

[0009] The assessment process for the impact of process parameters is refined to improve the accuracy of key parameter identification. The steps are as follows: A sample set is constructed within the range of values ​​for each process parameter, and multiple sets of parameter combination data are generated; Substitute the parameter combination into the calculation relationship between carbon emissions and processing time to obtain the corresponding result data; Variance decomposition was performed on the results data to calculate the first-order influence and overall influence of each parameter; The parameters are sorted according to their degree of influence, and the parameters with the greatest influence are selected as the targets for subsequent optimization.

[0010] The first-order influence and comprehensive influence of each process parameter are extracted from the variance decomposition results. A unified evaluation index is constructed, and the process parameters are sorted and divided according to the evaluation index to determine the set of key parameters. At the same time, a fixed value range is set for non-key parameters to reduce the parameter space and enhance the pertinence and stability of the subsequent optimization process.

[0011] The construction process of the multi-objective computational model is refined to improve its adaptability to actual production conditions. The steps are as follows: The carbon emissions of each process are summed up and calculated, and the overall carbon emissions are obtained by combining the process relationships. The processing time is accumulated or processed according to the process execution method to obtain the total processing time; Carbon emissions and processing time are standardized to form comparable target indicators; The operating range of the equipment, the allowable range of the process, and the cost constraints are incorporated into the calculation process to form a complete optimization calculation expression.

[0012] The following steps refine the dual-population collaborative optimization process to improve the solution capability under complex constraints: Individuals in the parameter space are initialized and divided into two populations with different evolutionary directions; Individuals that violate the constraints during the evolutionary process are evaluated, and their fitness is adjusted using a penalty coefficient. Based on the current population status, the penalty coefficient is dynamically adjusted using fuzzy rules to control the search range; Superior individuals are exchanged between two populations, and the individual set is continuously updated to obtain a parameter combination that meets the conditions.

[0013] Different fitness evaluation methods are set during the population division process. For populations facing constraints, individuals that meet the constraints are selected first. For populations facing goal optimization, the evaluation results of the objective function are strengthened. In the process of individual exchange, a selection mechanism is introduced to ensure that the exchanged individuals meet both the constraints and the goal performance, thereby improving the stability and effectiveness of parameter combinations.

[0014] The following steps refine the process of selecting and applying parameter combinations to improve the adaptability of optimization results in actual production: The obtained parameter combinations were categorized and organized, and the carbon emissions and processing time results corresponding to each combination were statistically analyzed. Parameter combinations are screened based on production needs, and the target parameter combination is determined; The target parameter combination is applied to the parameter settings of each process and then production is executed. The execution results are recorded and compared to evaluate the optimization effect and generate feedback data.

[0015] Limitations are made to ensure the stability and consistency of the parameter combination screening results. The consistency of the selected target parameter combinations is verified, and the parameter value range is corrected by combining historical operation data. The corrected parameter combinations are applied to the control of each process. At the same time, energy consumption and processing time during the execution process are collected in real time, and the collected results are compared with the calculation results in the screening stage to improve the reliability of parameter application and the consistency of optimization effect.

[0016] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention unifies and integrates the previously fragmented carbon emission calculations of various processes by constructing a multi-level mapping relationship between demand, manufacturing process, process parameters, and carbon emissions. This systematically expresses the correlation between different processes, thereby avoiding the problem of increased overall carbon emissions caused by optimizing a single process and achieving low-carbon collaborative optimization across the entire process. Simultaneously, by introducing the Sobol variance decomposition method, the influence of each process parameter on carbon emissions and processing time is quantitatively analyzed. This identifies key parameters, provides a clear basis for subsequent optimization, reduces the participation of invalid parameters, and improves the targeting and rationality of parameter selection.

[0017] This invention combines a dual-population collaborative strategy and a fuzzy logic adjustment mechanism in the optimization process, enabling constraint satisfaction and objective optimization to be handled separately yet collaboratively, thereby improving the feasibility and stability of parameter combinations under complex constraints. By dynamically adjusting the penalty coefficient, the search process is guided to achieve a reasonable balance at different stages, thus obtaining parameter combinations that meet multiple objective requirements. The resulting optimization results can be directly applied to existing manufacturing equipment; by adjusting process parameters, carbon emissions can be reduced, processing time shortened, and energy costs lowered. The overall solution requires no additional equipment investment and has good applicability and promotional value. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0019] Figure 1 This is a schematic diagram of the structure of the DMPC hierarchical mapping model of the present invention.

[0020] Figure 2 To create a carbon emission accounting boundary map for this invention.

[0021] Figure 3 This is a flowchart of the PFLCA optimization algorithm of the present invention.

[0022] Figure 4 This is a diagram showing the node relationships in the manufacturing process of the tubular heat exchanger of the present invention.

[0023] Figure 5 This is a comparison chart of the elasticity coefficients between carbon emissions and processing time for each process in this invention.

[0024] Figure 6 This is a comparison diagram of the Pareto frontiers of each process in this invention.

[0025] Figure 7 This is a comparison chart of the convergence curves of the objective function when using NSGA-III for solving.

[0026] Figure 8 This is a comparison chart of the convergence curves of the objective function when using IMOPSO to solve the problem.

[0027] Figure 9 This is a comparison chart of the convergence curves of the objective function when using GA to solve the problem.

[0028] Figure 10 This is a comparison chart of the convergence curves of the objective function when using PFLCA to solve the problem.

[0029] Figure 11This is a comparison table of process parameters and performance indicators before and after optimization using the method of this invention.

[0030] Figure 12 This is a comparison table of energy consumption and energy costs before and after optimization using the method of this invention. Detailed Implementation

[0031] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0032] This invention provides, for example Figures 1-12 The method for optimizing the low-carbon manufacturing process parameters of a tubular heat exchanger is shown below, with the specific steps as follows: The manufacturing process of tubular heat exchangers is systematically reviewed and structurally represented. The overall manufacturing process is divided into several typical steps according to the actual production sequence, including but not limited to key processing steps such as cutting, milling, rounding, and welding. For each of these steps, based on the carbon emission generation mechanism, they are further abstracted into multiple process objects directly related to carbon emissions. Each process object corresponds to specific processing behaviors and corresponding energy and material consumption processes, thus providing a unified analytical basis for subsequent carbon emission analysis.

[0033] Building upon the process breakdown, a multi-level correspondence is further established between demand, manufacturing process, process parameters, and carbon emissions. The demand side primarily reflects the product's functional indicators and quality requirements, such as dimensional accuracy, surface quality, and structural integrity. The manufacturing process side describes the complete processing path from raw materials to finished product. The process parameters side characterizes the key control variables involved in each process during actual execution, including processing speed, power, rotational speed, and current. The carbon emissions side represents the carbon emissions generated by energy consumption and material usage during each process. By mapping these different levels of relationships layer by layer, the previously scattered manufacturing information is unified and integrated, thus achieving a complete path expression: demand drives manufacturing, the manufacturing process maps to specific process parameters, and finally, carbon emissions are linked to the results.

[0034] Furthermore, based on the aforementioned multi-level relationships, the logical relationships between each process are formally described. For processes that must be executed sequentially in a fixed order, AND logic is used to express the dependencies between processes; for processes that can be performed in parallel or substituted, OR logic is used to represent the selection or parallel relationships between different process paths. By introducing AND / OR logic to uniformly express process relationships, the combination and execution order of different processes in the manufacturing process can be clearly reflected, making the complex manufacturing process computable and analyzable.

[0035] Under the constraints of the above logical relationships, the various process objects are connected in a predetermined order or combination to form a complete manufacturing process expression path. For processes executed sequentially, their processing time and carbon emissions are calculated in a progressively cumulative manner; for processes executed in parallel, the maximum processing time is taken based on the actual production cycle and resource allocation, while the carbon emissions are calculated by adding them together. In this way, consistent calculation of time and carbon emissions for different types of process combinations can be achieved within a unified framework.

[0036] Furthermore, the potential mutual influences between different processes are considered when constructing the aforementioned relationships. For example, the processing quality or parameter selection of a preceding process may affect the processing difficulty, processing time, and energy consumption level of subsequent processes. By introducing inter-process dependencies into the relationship expression, this cross-process influence can be further reflected, thereby avoiding the bias caused by analyzing a single process independently.

[0037] The above method transforms the originally discrete and complex manufacturing process into a set of operations with clearly defined relationships. It establishes a clear correspondence between demand, manufacturing processes, process parameters, and carbon emissions. Furthermore, it uses AND / OR logic to describe the execution relationships between operations, ultimately forming a unified expression foundation that can be used for subsequent carbon emission calculations and processing time assessments. This expression not only accurately reflects the sources and distribution of carbon emissions during multi-stage manufacturing but also provides a structured data foundation for subsequent process parameter optimization, facilitating the overall analysis and optimization of the manufacturing process.

[0038] Based on the breakdown of the manufacturing process and the expression of the relationships between procedures, a quantitative description is further provided on the relationship between the process parameters involved in each procedure and carbon emissions. Specifically, focusing on each procedure, a calculation relationship is established with process parameters as input and carbon emissions as output, taking into account both energy consumption and material consumption, thereby achieving a quantitative analysis of the sources of carbon emissions in the manufacturing process.

[0039] Regarding energy consumption, carbon emissions from different processes mainly originate from the electrical energy or other energy sources consumed during equipment operation. For each process, based on the equipment's operating characteristics, energy consumption can be expressed as a function of process parameters and processing time. For example, in the cutting process, the equipment's power level directly determines the energy consumption per unit time, while the processing time is related to the cutting speed and processing path length; in the milling process, parameters such as spindle speed and feed rate jointly affect cutting power and processing time; in the rounding process, roller speed and loading conditions affect equipment operating time and power requirements; in the welding process, welding current, voltage, and welding speed determine the heat input level and energy consumption. By analyzing the operating mechanisms of the above-mentioned processes, corresponding energy consumption expressions can be established for each.

[0040] Regarding material consumption, some processes involve material loss or the use of auxiliary materials, such as waste loss during cutting, welding materials, and shielding gases used in welding. These materials generate carbon emissions during production, transportation, and use, and therefore need to be included in the overall carbon emission calculation. For different types of material consumption, a carbon emission factor per unit mass of material is introduced to convert the material usage into corresponding carbon emissions, which are then added to the carbon emissions generated by energy consumption.

[0041] Building upon the above, the carbon emissions of a single process are expressed as the sum of energy consumption carbon emissions and material consumption carbon emissions. Energy consumption carbon emissions are obtained by multiplying the energy consumption amount by the energy carbon emission factor, while material consumption carbon emissions are obtained by multiplying the material usage amount by the corresponding carbon emission factor. This approach allows carbon emissions from different sources to be uniformly incorporated into a single calculation framework, achieving a complete description of the carbon emissions of a single process.

[0042] Furthermore, specific relationships between process parameters and carbon emissions are established for different types of processes. In the cutting process, cutting power and speed are used as inputs, and processing time is calculated based on the processing path length to obtain energy consumption, which is then converted into carbon emissions. In the milling process, spindle speed, feed rate, and depth of cut are used as inputs, and power requirements and processing time are estimated using a cutting force model or empirical formula to calculate carbon emissions. In the rounding process, roller speed and action time are used as the main variables to estimate equipment operating energy consumption and convert it into carbon emissions. In the welding process, welding current, voltage, and welding speed are used as core parameters to calculate welding heat input and duration, and the total carbon emissions are obtained by combining welding material consumption. By modeling different processes separately, the impact of various process parameters on carbon emissions can be reflected more accurately.

[0043] To improve the accuracy and applicability of the model, after establishing the aforementioned calculation relationships, actual production data is introduced to correct the model parameters. Specifically, power data, processing time data, and material consumption data during equipment operation are collected and compared with theoretical calculation results. The correlation coefficients in the model are then adjusted to make the calculation results closer to reality. This process involves fitting multiple sets of sample data to progressively optimize the model parameters, thereby improving the model's adaptability under different production conditions.

[0044] In addition, during the parameter correction process, the impact of different equipment states and operating conditions on energy consumption can be considered, such as equipment aging, load changes and environmental factors. Through statistical analysis of historical data, the model can be further corrected so that it can reflect the changes in the actual production environment.

[0045] The above method establishes a quantitative calculation relationship driven by process parameters, with energy and material consumption as intermediate variables, and carbon emissions as the output. This calculation relationship enables a refined description of carbon emissions from different processes and provides fundamental data support for subsequent multi-process carbon emission accumulation and optimization analysis. Furthermore, this method avoids the problem of relying solely on experience or statistical methods for estimation, giving carbon emission calculations clear physical meaning and traceability, thus improving the reliability and practicality of the overall optimization process.

[0046] Having established the quantitative relationship between process parameters and carbon emissions, it is necessary to identify and screen key process parameters affecting carbon emissions and processing time to improve the efficiency and reliability of subsequent optimization processes. Since the manufacturing process involves many types of process parameters, and there may be complex nonlinear relationships and coupling effects between different parameters, directly optimizing all parameters would not only significantly increase computational complexity but may also lead to slow convergence or even getting trapped in local solutions. Therefore, it is necessary to quantitatively assess the importance of each parameter before optimization to screen out the parameters that have the main impact on the target result.

[0047] To achieve the above objectives, the Sobol variance decomposition method is introduced to globally analyze the influence of each process parameter. Based on the principle of variance decomposition, this method decomposes the uncertainty of the model output into the contributions of each input parameter and their interactions, thereby enabling a global assessment of the magnitude of each parameter's influence on the output. Specifically, based on the established relationship between process parameters and carbon emission calculation, each process parameter is treated as an independent input variable. Random sampling is performed within its defined range to construct a sample set covering the parameter space, and the corresponding total carbon emissions and total processing time are calculated.

[0048] Based on this, the variance of the model output is decomposed, and the first-order sensitivity and total-order sensitivity of each process parameter to total carbon emissions and total processing time are calculated separately. The first-order sensitivity characterizes the degree of influence of independent changes in a single parameter on the output, while the total-order sensitivity reflects the combined effect of the parameter considering its interaction with other parameters. By comparing the first-order and total-order sensitivities, significant coupling relationships between parameters can be further identified.

[0049] After obtaining the sensitivity results for each process parameter, the parameters are sorted according to their numerical values ​​to identify the key parameters that have a major impact on total carbon emissions and the key parameters that have a major impact on total processing time. Parameters that have high sensitivity in both objectives can be prioritized for optimization; for parameters with low sensitivity, their search range can be appropriately reduced or kept at a fixed value in subsequent optimization processes to reduce optimization dimensions and improve solution efficiency.

[0050] Furthermore, to apply the sensitivity analysis results to subsequent optimization processes, the sensitivity of each process parameter is normalized to obtain corresponding weighting coefficients. Specifically, based on the first-order or total-order sensitivity of each parameter, they can be mapped to a unified numerical range, and the relative weight of each parameter can be calculated through proportional relationships. These weighting coefficients can reflect the relative importance of different parameters in influencing carbon emissions and processing time.

[0051] In multi-stage manufacturing processes, the types of process parameters involved in different stages and their degree of influence vary. Therefore, when constructing weighting coefficients, the parameter weights can be categorized and organized according to the stage division results, thereby obtaining the distribution of parameter importance within and between stages. This approach more clearly reflects the contribution of different stages to overall carbon emissions and processing time, providing a reference for subsequent optimization.

[0052] Furthermore, during sensitivity analysis, multiple sampling and repeated calculations can improve the stability of the results and reduce the impact of random errors. Simultaneously, the number of samples and the parameter range can be adjusted according to actual needs to achieve a balance between computational accuracy and computational cost.

[0053] Through the above analysis, a method for evaluating the impact of process parameters based on Sobol variance decomposition is developed, enabling the effective identification and screening of key parameters in multi-stage manufacturing processes. This method can perform a comprehensive analysis of the parameter space without relying on local gradient information, thus avoiding the bias problems that may arise from local analysis methods.

[0054] Finally, by normalizing the sensitivity results, the weight coefficients corresponding to each process parameter are obtained and used as important reference information in subsequent optimization processes. During the optimization process, these weights can be used to guide the search direction or adjustment range of different parameters, thereby improving optimization efficiency and enhancing the stability and rationality of the optimization results. Simultaneously, this method effectively reduces the dimensionality of the problem by screening parameters before optimization, providing a solid foundation for multi-objective optimization under complex constraints.

[0055] Based on the identification and weight determination of key process parameters, a computational model for multi-objective collaborative optimization is further constructed to achieve comprehensive control of carbon emissions and processing efficiency during the manufacturing process. Considering that the manufacturing process of tubular heat exchangers requires both reducing carbon emissions and meeting production cycle requirements, total carbon emissions and total processing time are selected as the main optimization objectives. By establishing a multi-objective function, the simultaneous constraint and optimization of these two types of indicators can be achieved.

[0056] Specifically, based on the established relationships between process parameters and carbon emission calculations, as well as processing time calculations, the carbon emissions of each process are accumulated according to a predetermined process sequence or combination to obtain the total carbon emissions of the entire manufacturing process. Simultaneously, according to the processing time calculation method for each process, sequential processes are accumulated step-by-step, while parallel processes are processed by taking the maximum value according to the processing cycle time, thus obtaining the total processing time of the entire manufacturing process. Through this method, the dispersed information in multi-process manufacturing can be transformed into a unified overall evaluation index.

[0057] In constructing the objective function, to avoid the influence of different units, total carbon emissions and total processing time are normalized to ensure they fall within a uniform numerical range, thus facilitating subsequent optimization calculations. Normalization can be performed by linearly transforming each indicator based on its maximum and minimum values ​​within historical data or a preset range, making different objectives comparable during the optimization process.

[0058] Building upon this, the parameter weights obtained from sensitivity analysis are incorporated into the objective function construction. By assigning different weight coefficients to different process parameters, the optimization process focuses more on key parameters that significantly impact the target result. Specifically, for parameters with high sensitivity, their weight is increased during the objective function calculation, thus prioritizing adjustments to these parameters during the search process; for parameters with low sensitivity, their influence is reduced to minimize ineffective searches. In this way, parameter importance information is introduced into the optimization process, making the search more targeted and improving optimization efficiency.

[0059] Regarding constraints, considering various limiting factors in actual production, the model incorporates equipment rated capacity, process allowable range, and processing cost range as constraints. Equipment rated capacity mainly includes operating limits such as maximum power and maximum speed, ensuring the optimization results remain within the equipment's tolerance range. The process allowable range defines the value intervals of each process parameter to guarantee processing quality and safety. The processing cost range controls overall manufacturing costs, preventing excessive costs while reducing carbon emissions or shortening time. By introducing these multiple constraints, the optimization results can more closely reflect the actual production environment.

[0060] In terms of model representation, the optimization problem is expressed as solving a multi-objective function while satisfying the aforementioned constraints. Since there may be conflicting relationships between the objective functions—for example, reducing carbon emissions might lead to increased processing time—the optimization process no longer pursues a single optimal solution, but rather seeks a set of feasible solutions that achieve a balance among multiple objectives. In this way, multiple alternative solutions can be provided to meet different production needs.

[0061] Furthermore, when constructing the optimization model, the objective function can be extended according to actual needs. For example, processing cost can be introduced as a third objective, or different objectives can be prioritized to adapt to different application scenarios. In the process of weighing multiple objective functions, the emphasis on different objectives can be flexibly adjusted by adjusting the weights or adopting other evaluation methods.

[0062] Through the above steps, a multi-objective optimization computational framework incorporating objective functions and constraints is formed. Parameter sensitivity information is integrated into the objective function, ensuring that the optimization process considers both the overall objective and the influence of key parameters. This method not only improves the efficiency of optimization computation but also obtains parameter combinations that better meet practical needs under complex constraints, providing a clear computational foundation for subsequent optimization solutions.

[0063] Example 1: Optimization of manufacturing process for tubular heat exchangers.

[0064] Taking a tubular heat exchanger manufactured by a chemical equipment company as an example, its manufacturing process includes: laser cutting of steel plates, milling of end faces, rolling forming, and submerged arc welding. The existing process parameters are set based on experience: laser cutting power 25kW, milling spindle speed 1900r / min, rolling speed 5.6r / min, and welding current 693A. The corresponding total carbon emissions are... The total processing time was 80.9 minutes.

[0065] 1. According to Figure 1The structure shown decomposes the manufacturing process into four carbon-related units: cutting, milling, rounding, and welding. According to the actual process logic, these four units are connected in series (AND). The process parameters corresponding to each unit are: laser power... Spindle speed P2, roller speed P3, welding current P4.

[0066] 2. Based on the principles of energy conservation and material balance, establish carbon emission models for each unit, as follows: In the formula, For laser power, For electro-optical conversion efficiency, For cutting time, It is a carbon emission factor for electricity.

[0067] Subscript , These represent spindle speed, machining power, lubricant consumption rate, lubricant density, and carbon strength, respectively. Depth of cut. ,width Number of teeth Z and feed per tooth These factors collectively determine the cutting amount V. The numerator of the equation reflects the instantaneous change in power with rotational speed, while dividing by n takes into account the shortened cycle time at high speeds. These terms together elucidate the trade-off between rotational speed and load prevalent in high-speed forming processes, which has a significant impact on energy consumption performance.

[0068] in This indicates the operating speed of the rolling machine. and These are lubrication rate and lubricant discharge coefficient, respectively. This is the total turning length. and These represent the optimal operating power and speed, respectively. Parameters and These are defined as machine tool efficiency and its loss coefficient, respectively. This configuration explicitly considers deviations from the rated operating point and auxiliary energy consumption related to lubrication, which conforms to accepted principles of machine tool energy consumption modeling.

[0069] in Indicates welding current. Indicates the specific heat capacity of a material. This indicates the mass of welding wire consumed. The temperature difference between ambient temperature and operating temperature is represented by... The power factor and arc thermal efficiency of the welding machine are respectively expressed as... and Representation. Coefficient and These correspond to the wire loss effect and the flux ratio effect, respectively. The constant term reflects the baseline implicit emission of the solder material, while the fractional term describes the nonlinear sensitivity of the arc efficiency to current deviation.

[0070] Total carbon emissions are Total processing time 3. Using the Saltelli sampling method, Sobol sensitivity analysis was performed on the four parameters within ±20%.

[0071] The calculated sensitivity (ST) of total carbon emissions to each parameter is as follows: 0.8402, 0.0935, 0.0268, The value is 0.0388. This indicates that cutting power has the absolute dominant impact on carbon emissions, followed by rounding, while milling and welding have negligible effects. The elasticity of processing time to each parameter is -1, but the absolute values ​​are not significantly different. The normalized weights of each process are calculated based on sensitivity. .

[0072] 4. Construct a multi-objective optimization model, whose optimization objective is: Its constraints are as follows: Device power limit: ; Cost constraint: Optimized energy consumption cost Initial energy consumption cost. Processing quality constraints (hidden within parameter boundaries).

[0073] To accelerate the search, a sensitivity-guided auxiliary target is used. However, the final optimization still uses the original physical target for non-dominated sorting.

[0074] 5. Set the population size to 200 and the maximum number of iterations to 200, and use the PFLCA algorithm proposed in this invention to solve the problem.

[0075] In comparison, the NSGA-III, IMOPSO, and GA algorithms were used simultaneously for solving the problem.

[0076] The convergence process of each algorithm is as follows: Figure 7As shown, PFLCA converged to a stable value in 16 generations, with the shortest calculation time (0.255s), and the final process parameter combination obtained was: P1=28.1kW, P2=1870r / min, P3=5.2r / min, P4=780.6A.

[0077] 6. The performance indicators before and after parameter optimization are compared as follows: Total carbon emissions compared to before optimization After optimization, it achieved The optimization rate was 3.29%. The total time was reduced from 80.9 hours before optimization to 70.3 hours after optimization, representing an optimization rate of 13.20%. Total carbon emissions increased from 4188.8 kWh before optimization to 4057.0 kWh after optimization, representing an optimization rate of 3.17%. The results show that, under strict process constraints, the method of this invention achieves multi-objective synergistic optimization of carbon emissions, processing time, and cost. The processing time is significantly reduced, attributable to the fact that processing time is sensitive to all parameters and has a relatively larger adjustment range. Although the optimization of carbon emissions is limited by the boundaries of the dominant parameters, it still achieves effective improvement, verifying the practicality of the method.

[0078] 7. The analysis results show that sensitivity analysis reveals that cutting power is the main contributor to carbon emissions, and its feasible adjustment range is limited by the rated power of the equipment and the cutting quality, thus limiting the optimization space for carbon emissions. In contrast, processing time is affected by multiple parameters, offering greater flexibility for adjustment and thus allowing for greater improvement. This invention, through a sensitivity-guided search strategy, concentrates computational resources on high-impact parameters, avoiding ineffective searches on low-sensitivity parameters, thereby obtaining a high-quality feasible solution within a finite number of iterations.

[0079] This invention unifies and integrates the previously scattered carbon emission calculations of various processes by constructing a multi-level mapping relationship between demand, manufacturing process, process parameters, and carbon emissions. This systematically expresses the correlation between different processes, thereby avoiding the problem of increased overall carbon emissions caused by optimizing a single process and achieving low-carbon collaborative optimization for the entire process. Simultaneously, by introducing the Sobol variance decomposition method, the influence of each process parameter on carbon emissions and processing time is quantitatively analyzed. This identifies key parameters and establishes weight relationships, providing a clear basis for subsequent optimization, reducing the participation of invalid parameters, and improving the targeting and rationality of parameter selection.

[0080] This invention combines a dual-population collaborative strategy and a fuzzy logic adjustment mechanism in the optimization process, enabling constraint satisfaction and objective optimization to be handled separately yet collaboratively, thereby improving the feasibility and stability of parameter combinations under complex constraints. By dynamically adjusting the penalty coefficient, the search process is guided to achieve a reasonable balance at different stages, thus obtaining parameter combinations that meet multiple objective requirements. The resulting optimization results can be directly applied to existing manufacturing equipment; by adjusting process parameters, carbon emissions can be reduced, processing time shortened, and energy costs lowered. The overall solution requires no additional equipment investment and has good applicability and promotional value.

[0081] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for optimizing low-carbon manufacturing process parameters of a tubular heat exchanger, characterized in that, Includes the following steps: The manufacturing process of tubular heat exchangers is broken down into multiple carbon emission-related process objects, and the correspondence between process parameters and carbon emissions is established. The AND / OR logic is used to describe the sequential and parallel relationships between each process, forming process association relationships. Based on the process relationships, for each process object, according to the energy consumption and material consumption, a calculation relationship is established with process parameters as input and carbon emissions as output, and the parameters are corrected in combination with production data; Based on the calculation relationship, the Sobol variance decomposition method is used to analyze the impact of each process parameter on the overall carbon emission and overall processing time, and to identify the main influencing parameters. Based on the main influencing parameters, a calculation model is established with overall carbon emissions and overall processing time as objectives, and the equipment capacity range, process allowable range and cost range are added as constraints. For the computational model, an optimization method involving two populations is introduced, one to satisfy the constraints and the other to optimize the objective. The penalty coefficient is dynamically adjusted through fuzzy logic and individuals are exchanged between populations to obtain the Letopa front solution set. Based on the solution set of the Letopascal front, target process parameters are selected in combination with production requirements and applied to the manufacturing process of tubular heat exchangers.

2. The method for optimizing low-carbon manufacturing process parameters of a tubular heat exchanger according to claim 1, characterized in that, The steps to refine the correlation between manufacturing processes and carbon emissions, and improve the accuracy and calculability of expressing multi-process relationships, are as follows: Each process is numbered according to the actual processing sequence, and the dependencies and parallel processing relationships between each process are clearly defined. Based on product quality requirements and manufacturing processes, the process parameters corresponding to different processes are classified and collected. The process parameters of each process are linked and recorded with the corresponding energy consumption and material consumption; By integrating process sequence, parameter information, and carbon emission relationships through a unified expression method, a relational expression result that can be used for calculation is formed.

3. The method for optimizing low-carbon manufacturing process parameters of a tubular heat exchanger according to claim 2, characterized in that, The following steps refine the calculation relationship between process parameters and carbon emissions to improve the accuracy of carbon emission calculations: The corresponding energy consumption form is determined based on the operating characteristics of each process, and the energy consumption generated during the processing is calculated. By combining energy carbon emission coefficients, energy consumption is converted into corresponding carbon emissions; The material losses and auxiliary material consumption generated during the processing are statistically analyzed, and the carbon emission coefficient of the materials is introduced for conversion. The carbon emissions from energy consumption and carbon emissions from material consumption are superimposed to obtain the carbon emission results for each process.

4. The method for optimizing low-carbon manufacturing process parameters of a tubular heat exchanger according to claim 3, characterized in that, The assessment process for the impact of process parameters is refined to improve the accuracy of key parameter identification. The steps are as follows: A sample set is constructed within the range of values ​​for each process parameter, and multiple sets of parameter combination data are generated; Substitute the parameter combination into the calculation relationship between carbon emissions and processing time to obtain the corresponding result data; Variance decomposition was performed on the results data to calculate the first-order influence and overall influence of each parameter; The parameters are sorted according to their degree of influence, and the parameters with the greatest influence are selected as the targets for subsequent optimization.

5. The method for optimizing low-carbon manufacturing process parameters of a tubular heat exchanger according to claim 4, characterized in that, The first-order influence and comprehensive influence of each process parameter are extracted from the variance decomposition results. A unified evaluation index is constructed, and the process parameters are sorted and divided according to the evaluation index to determine the set of key parameters. At the same time, a fixed value range is set for non-key parameters to reduce the parameter space and enhance the pertinence and stability of the subsequent optimization process.

6. The method for optimizing low-carbon manufacturing process parameters of a tubular heat exchanger according to claim 4, characterized in that, The construction process of the multi-objective computational model is refined to improve its adaptability to actual production conditions. The steps are as follows: The carbon emissions of each process are summed up and calculated, and the overall carbon emissions are obtained by combining the relationships between the processes. The processing time is accumulated or processed according to the process execution method to obtain the total processing time; Carbon emissions and processing time are standardized to form comparable target indicators; The operating range of the equipment, the allowable range of the process, and the cost constraints are incorporated into the calculation process to form a complete optimization calculation expression.

7. The method for optimizing low-carbon manufacturing process parameters of a tubular heat exchanger according to claim 6, characterized in that, The following steps refine the dual-population collaborative optimization process to improve the solution capability under complex constraints: Individuals in the parameter space are initialized and divided into two populations with different evolutionary directions; Individuals that violate the constraints during the evolutionary process are evaluated, and their fitness is adjusted using a penalty coefficient. Based on the current population status, the penalty coefficient is dynamically adjusted using fuzzy rules to control the search range; Superior individuals are exchanged between two populations, and the individual set is continuously updated to obtain a parameter combination that meets the conditions.

8. The method for optimizing low-carbon manufacturing process parameters of a tubular heat exchanger according to claim 7, characterized in that, Different fitness evaluation methods are set during the population division process. For populations facing constraints, individuals that meet the constraints are selected first. For populations facing goal optimization, the evaluation results of the objective function are strengthened. In the process of individual exchange, a selection mechanism is introduced to ensure that the exchanged individuals meet both the constraints and the goal performance, thereby improving the stability and effectiveness of parameter combinations.

9. The method for optimizing low-carbon manufacturing process parameters of a tubular heat exchanger according to claim 7, characterized in that, The following steps refine the process of selecting and applying parameter combinations to improve the adaptability of optimization results in actual production: The obtained parameter combinations were categorized and organized, and the carbon emissions and processing time results corresponding to each combination were statistically analyzed. Parameter combinations are screened based on production needs, and the target parameter combination is determined; The target parameter combination is applied to the parameter settings of each process and then production is executed. The execution results are recorded and compared to evaluate the optimization effect and generate feedback data.

10. The method for optimizing low-carbon manufacturing process parameters of a tubular heat exchanger according to claim 9, characterized in that, Limitations are made to ensure the stability and consistency of the parameter combination screening results. The consistency of the selected target parameter combinations is verified, and the parameter value range is corrected by combining historical operation data. The corrected parameter combinations are applied to the control of each process. At the same time, energy consumption and processing time during the execution process are collected in real time, and the collected results are compared with the calculation results in the screening stage to improve the reliability of parameter application and the consistency of optimization effect.

Images