A method and system for collaboratively controlling thermal processing quality and energy consumption of large aerospace components

Abstract

The present invention belongs to the field of launchers and discloses a method and system for coordinated control of the quality and energy consumption of hot-processing of large aerospace components. The method includes: collecting data to form an equipment energy consumption characteristic data set; setting multiple sets of multi-pass process parameters and simulating and calculating corresponding quality data to form a quality characteristic data set; inputting the energy consumption characteristic data set and the quality characteristic data set into an adaptive fuzzy neural network to be trained, and training them until the trained models converge, thereby obtaining an energy consumption prediction model and a quality prediction model; inputting data into the two models to predict the standard value of the energy consumption required for processing the component, and obtaining a quality prediction value for the hot-processed component product; and adjusting and optimizing the process of subsequent processes using a particle swarm algorithm to achieve coordinated control of the quality and energy consumption of the hot-processed component. With the help of the present invention, flexible control of the entire forging process of large aerospace components can be achieved, stabilizing product quality and energy consumption, and improving factory production efficiency.

Description

Technical Field

[0001] The present invention belongs to but is not limited to the field of launcher technology, and in particular relates to a method and system for coordinated control of thermal processing quality and energy consumption of large aerospace components. Background Art

[0002] Large aerospace components, such as aircraft landing gear and rocket engine heads, require extremely high strength and toughness. They are typically forged using large-tonnage forging presses. The complete manufacturing process includes billet making, die forging, and heat treatment, with billet making and die forging involving heating, forging, and heat preservation. The forgings undergo repeated heating and cooling, as well as extensive deformation, during the manufacturing process, resulting in a complex process. Furthermore, the bulky processing equipment makes it difficult to maintain stable production processes, which can easily lead to fluctuations and affect the final quality of the forgings.

[0003] The manufacturing of large forgings is also a highly energy-intensive process. Currently, forging plants lack management over their production energy consumption, making it difficult to proactively and effectively control energy consumption.

[0004] Quality and cost are the primary criteria for determining the quality of a factory's products, with energy costs accounting for a significant portion of forging costs. Due to the significant fluctuations in quality and energy consumption during the current forging process for large aerospace components, substandard quality often leads to additional rework and repair, further increasing costs.

[0005] In view of the above analysis, the technical problems that need to be solved urgently in the existing technology are:

[0006] The factory urgently needs to dynamically control the entire process of thermal processing of large components to ensure the stability of product quality and energy consumption.

[0007] Chinese patent CN115358889A discloses a method for controlling energy consumption of industrial devices based on a back propagation (BP) neural network. The method comprises: collecting historical energy consumption information and influencing factor information of industrial devices and their corresponding historical data to form an energy consumption feature dataset; inputting the energy consumption feature dataset into a BP neural network model to be trained until the trained BP neural network model converges, thereby obtaining a trained BP neural network model; inputting the collected real-time energy consumption data of industrial devices into the trained BP neural network model to run the model, thereby obtaining a predicted energy consumption value of the industrial device; and when the absolute value of the deviation between the predicted energy consumption value and the actual energy consumption value exceeds a set deviation fluctuation value, the industrial device issues an abnormal alarm. This method can use the industrial device energy consumption prediction model to alarm abnormal energy consumption performance in production, but it cannot automatically adjust the equipment to return energy consumption to normal, and its intelligence level is insufficient.

[0008] Chinese patent CN118034212A discloses a forging process optimization system and optimization method based on digital twins, including: a physical entity acquisition module and a digital twin system, wherein the physical entity acquisition module includes a data acquisition module, and the data acquisition module is used to collect information data during the operation of the physical entity acquisition module. This patent integrates digital twin technology into the forging process, and guides the process optimization process by comparing the performance of the metal after actual forging with the performance after forging simulated by the digital twin. The above optimization method updates the process parameters based on the quality of the metal after processing, and cannot adjust the process during production to control the final quality. Due to the large fluctuations in the production process of large aerospace forgings, the quality of each product also fluctuates greatly. At the same time, if the quality does not meet the standards, it will face high rework or scrap costs. The above method cannot dynamically adjust the production of large forgings in a highly volatile production environment. Summary of the Invention

[0009] In response to the problems existing in the prior art, the present invention provides a method and system for coordinated control of thermal processing quality and energy consumption of large aerospace components.

[0010] The present invention is achieved by providing a method for collaboratively controlling the thermal processing quality and energy consumption of large aerospace components, characterized in that the method specifically comprises:

[0011] S1: Collect energy consumption data and process data of each device during the thermal processing of aerospace components, and combine them with existing historical energy consumption data to form an equipment energy consumption characteristic data set;

[0012] S2: Import the aerospace component model and mold model into the finite element simulation software. Based on the actual hot working process, set multiple sets of multi-pass process parameters for simulation calculation to obtain the corresponding quality data. Combine the quality inspection data of each batch of components and the corresponding process data to form a quality feature data set;

[0013] S3: Inputting the energy consumption characteristic data set and the quality characteristic data set into the adaptive fuzzy neural network to be trained respectively until the trained model converges, thereby obtaining an energy consumption prediction model for each hot processing equipment and a multi-pass quality prediction model for hot processing components;

[0014] S4: Inputting the preset thermal processing process data into the energy consumption prediction model to obtain the standard energy consumption prediction value of each device during the thermal processing of aerospace components. Based on the standard thermal processing process route data, the standard value of energy consumption required for processing the component is predicted as a reference value for energy consumption in actual processing;

[0015] S5: Input the real-time process data in actual hot working production into the quality prediction model to obtain the quality prediction value of the hot working component product;

[0016] S6: Based on whether the quality prediction value meets the quality requirements and the deviation between the actual equipment energy consumption value and the standard energy consumption prediction value, the particle swarm algorithm is used to adjust and optimize the process of subsequent processes to achieve coordinated control of the quality and energy consumption of hot-processed components.

[0017] Furthermore, the S1, hot working process includes blank making, die forging and heat treatment; the energy consumption data and process data of each equipment include the input power, pressing speed, pressing force, die temperature, component temperature of the forging press, the input power or gas consumption rate, heating rate, and temperature change curve in the furnace of the heating furnace or heat treatment furnace.

[0018] Furthermore, the S2 quality data includes the grain size, yield strength, tensile strength, etc. of the final product of the hot-processed component; the process data includes the forging rate of multiple forging passes, component temperature before forging, mold temperature, heat treatment heating rate, holding temperature, holding time, etc.

[0019] Furthermore, the energy consumption prediction model in S3 is established for different hot working equipment, including: open forging presses, die forging presses, pre-forging heating furnaces and heat treatment furnaces. The input of the model is the time series of equipment process data, and the output is the time series of equipment power / gas consumption rate. After processing, it can be converted into the total power consumption / gas consumption of the equipment over a period of time. The prediction model can simulate and calculate the corresponding input power or gas consumption rate when the equipment is operated according to a series of set process parameters in actual production;

[0020] Quality prediction models are established for different output targets, including multiple prediction models with quality indicators such as grain size, yield strength, and tensile strength as output values. The model input is the full-process process parameter matrix. The model establishes the relationship between the final quality of the forging and the complete hot working process technology.

[0021] Furthermore, in S6, when it is predicted that the final quality of the component is unqualified, a particle swarm algorithm is used to perform multi-objective optimization on the subsequent process. The process parameters are updated through the algorithm and input into the quality and energy consumption prediction model. The process parameters are updated with the predicted final quality of the component as a constraint and the predicted minimum required energy consumption as the goal. Finally, optimization is achieved. The objective function can be expressed as:

[0022]

[0023] Among them, X represents the multidimensional process matrix composed of process parameters, Q i (X) indicates the predicted values ​​of different qualities, a i ~b i It represents the required range of different qualities, and E(X) represents the total energy consumption of the whole process, which is the sum of the energy consumption forecasts of each equipment.

[0024] Another object of the present invention is to provide a large-scale aerospace component thermal processing quality and energy consumption coordinated control system, the system specifically comprising:

[0025] Data acquisition module, which collects and processes data to build equipment energy consumption characteristic data sets and quality characteristic data sets;

[0026] Energy consumption prediction module, based on the energy consumption prediction model, obtains the standard energy consumption prediction value of each equipment in the thermal processing of aerospace components;

[0027] The quality prediction module obtains the quality prediction value of the hot-processed component product based on the quality prediction model;

[0028] The optimization module uses particle swarm optimization to perform multi-objective optimization on subsequent processes.

[0029] In combination with the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solutions to be protected by the present invention are as follows:

[0030] First, the present invention establishes an energy consumption prediction model for the equipment through the operation data and energy consumption data of the equipment in actual production. The model can predict the energy consumption of the equipment under different processes. Through multiple equipment energy consumption models, the energy consumed in the entire product processing process can be predicted.

[0031] The present invention integrates process data and quality inspection results from actual production with simulation process data and performance results based on the material property evolution mechanism to establish a quality prediction model for finished forgings. This model can predict the final forging quality based on the process parameters of each processing stage.

[0032] The present invention can achieve quality and energy consumption management of the entire hot processing process with the help of quality and energy consumption prediction models. By comparing the energy consumption prediction results with the equipment energy consumption monitored during actual processing, the energy consumption fluctuation situation can be judged. By comparing the quality prediction results with the required quality, the quality fluctuation situation can be judged. Based on the energy consumption and quality fluctuation situation, the subsequent processes are combined and optimized with the goal of achieving quality standards and minimizing energy consumption.

[0033] With the help of the present invention, flexible control of the entire forging process of large-scale aerospace components can be achieved, product quality and energy consumption can be stabilized, and factory production efficiency can be improved.

[0034] Second, the present invention provides a system for the coordinated control of the quality and energy consumption of thermal processing of large aerospace components. By integrating modules such as data acquisition, finite element simulation, adaptive fuzzy neural network training, energy consumption and quality prediction, and optimization control, it achieves dual optimization of quality control and energy consumption management during the thermal processing process.

[0035] The data acquisition unit collects energy consumption data and process parameters of each thermal processing equipment in real time through sensors and data interfaces, and combines it with historical data to form a comprehensive equipment energy consumption characteristic data set.

[0036] The finite element simulation module uses finite element analysis software to simulate and calculate multiple sets of process parameters for aerospace components and mold models, generate corresponding quality data, and combine it with actual quality inspection data to construct a quality feature data set.

[0037] The adaptive fuzzy neural network training module generates energy consumption prediction models and quality prediction models by training the adaptive fuzzy neural network, ensuring high accuracy and reliability of the prediction results.

[0038] The energy consumption prediction module generates standard energy consumption prediction values ​​for each device based on preset process parameters and uses energy consumption prediction models to provide a reference benchmark for energy consumption management in actual production.

[0039] The quality prediction module inputs the quality prediction model through real-time process data to generate the quality prediction value of the component to ensure quality control during the hot working process.

[0040] The optimization control unit comprehensively considers the deviation between quality and energy consumption, and dynamically optimizes and adjusts the process parameters through the particle swarm algorithm to achieve coordinated control of quality and energy consumption during the hot working process, thereby improving production efficiency and economic benefits.

[0041] The system for coordinated control of thermal processing quality and energy consumption of large aerospace components of the present invention has the following significant technical effects in industrial applications:

[0042] 1. Achieve dual optimization of quality and energy consumption: By integrating quality prediction and energy consumption prediction models, the system can optimize process parameters, reduce energy consumption and improve production efficiency while ensuring component quality.

[0043] 2. Improve prediction accuracy and reliability: Adopting an adaptive fuzzy neural network for model training, combined with finite element simulation and historical data, significantly improves the accuracy of energy consumption and quality predictions and reduces prediction errors.

[0044] 3. Realize real-time monitoring and dynamic adjustment: The system can monitor the process parameters and equipment energy consumption during the hot working process in real time, and realize real-time optimization of the production process by optimizing the dynamic adjustment of the control unit to ensure the efficiency and stability of the production process.

[0045] 4. Reduce production costs and energy consumption: By optimizing process parameters, the system effectively reduces energy consumption during the thermal processing process, reduces production costs, improves resource utilization, and meets the requirements of green production and sustainable development.

[0046] 5. Enhance the intelligence and automation level of the system: Integrate multiple intelligent modules to realize the automation of the entire process from data collection, prediction to optimization control, reduce manual intervention, and improve the intelligence level and operational efficiency of the system.

[0047] 6. Adapt to the complex processing requirements of large aerospace components: The system is designed based on the thermal processing characteristics of large aerospace components, and can handle complex process requirements with multiple passes and multiple parameters, ensuring the quality and energy consumption management of components during high-strength and high-precision processing.

[0048] Through the above technical solutions, the present invention effectively solves the problem of separation between quality control and energy consumption management in the existing thermal processing of large aerospace components, realizes the coordinated optimization of the two, significantly improves the overall efficiency and economic benefits of the thermal processing process, and has broad industrial application prospects and significant technological progress.

[0049] The expected benefits and commercial value of this invention after transformation are as follows: This invention will ensure stable production quality and energy consumption for large aerospace forgings, helping to reduce production costs and generating significant economic benefits. The coordinated control system for aerospace hot-working component quality and energy consumption provided by this invention will promote the transformation and upgrading of the traditional aerospace hot-working industry toward high-quality, high-efficiency, and low-consumption production.

[0050] Third, the method for collaboratively controlling the quality and energy consumption of thermal processing of large aerospace components provided by the present invention effectively solves several major technical problems in the prior art by dually predicting and controlling energy consumption and quality, and achieves significant technological progress:

[0051] 1. Accurately control energy consumption and reduce waste

[0052] Prior art lacks precise prediction and optimization methods for controlling equipment energy consumption during hot working processes, often leading to energy waste. This invention collects equipment energy consumption characteristic data and combines it with an adaptive fuzzy neural network model and particle swarm optimization algorithm to achieve real-time prediction and dynamic adjustment of energy consumption, significantly reducing energy consumption fluctuations, eliminating unnecessary energy consumption, and significantly improving energy conservation.

[0053] 2. Improve the stability and consistency of component quality

[0054] In traditional hot working, quality instability often occurs due to imprecise process parameter control. This invention, through a quality prediction model, enables real-time monitoring of product quality and timely adjustment of subsequent process parameters when deviations are detected. This real-time quality optimization and control ensures the stability and consistency of component quality, reduces product rejection rates caused by quality fluctuations, and improves processing quality.

[0055] 3. Reduce production costs and improve economic benefits

[0056] This invention optimizes thermal processing parameters through standardized energy consumption and quality prediction models, reducing unnecessary energy consumption and material waste, directly lowering production costs. Furthermore, the reduced energy consumption and optimized process parameters shorten processing time and improve equipment utilization, bringing significant economic benefits to the enterprise.

[0057] 4.Automation and intelligent control of process parameters

[0058] The solution of the present invention relies on adaptive fuzzy neural networks and particle swarm algorithms to achieve intelligent control of process parameters, transforming the heat treatment process from relying on manual experience to relying on intelligent system control. It is convenient to operate and reduces human errors, making the production process more efficient and controllable.

[0059] 5. Improve the adsorption performance and other key performance of hot-processed components

[0060] By optimizing processes such as oxidation baking, the present invention further improves key physical properties such as the adsorption performance of the component, meeting the strict requirements of high-precision aerospace equipment, thereby expanding the application range of the product and enhancing the competitiveness of the product in the aerospace field.

[0061] 6. Promote efficient use of resources and environmentally friendly production

[0062] This invention achieves resourceful utilization of waste gas and optimized energy consumption management, aligning with the development trend of energy conservation and emission reduction, promoting environmentally friendly production, and demonstrating the advantages of sustainable development. In aerospace component manufacturing, this environmentally friendly and efficient solution significantly improves the sustainability of the production process.

[0063] Through the above technical solution, the present invention effectively solves the problems of energy waste, unstable quality, high production costs and so on in the prior art, and achieves significant technological progress. It not only promotes the upgrading of aerospace component production technology, but also enhances the market competitiveness of the products. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 This is a flow chart of a method for collaboratively controlling the quality and energy consumption of thermal processing of large aerospace components provided by an embodiment of the present invention;

[0065] Figure 2 Schematic diagram of the adaptive fuzzy neural network structure provided by an embodiment of the present invention;

[0066] Figure 3 It is a module diagram for coordinated control of thermal processing quality and energy consumption of large aerospace components provided by an embodiment of the present invention.

[0067] Figure 4 This is the hot working flow chart of the die forging.

[0068] Figure 5 A plot of the iterative optimization prediction results for the average grain size.

[0069] Figure 6 Graph showing the iterative optimization prediction results for total energy consumption. DETAILED DESCRIPTION

[0070] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0071] like Figure 1 As shown, an embodiment of the present invention provides a method for collaboratively controlling the thermal processing quality and energy consumption of large aerospace components, the method specifically comprising:

[0072] S1: Collect energy consumption data and process data of each device during the thermal processing of aerospace components, and combine them with existing historical energy consumption data to form an equipment energy consumption characteristic data set;

[0073] S2: Import the aerospace component model and mold model into the finite element simulation software. Based on the actual hot working process, set multiple sets of multi-pass process parameters for simulation calculation to obtain the corresponding quality data. Combine the quality inspection data of each batch of components and the corresponding process data to form a quality feature data set;

[0074] S3: Inputting the energy consumption characteristic data set and the quality characteristic data set into the adaptive fuzzy neural network to be trained respectively until the trained model converges, thereby obtaining an energy consumption prediction model for each hot processing equipment and a multi-pass quality prediction model for hot processing components;

[0075] S4: Inputting the preset thermal processing process data into the energy consumption prediction model to obtain the standard energy consumption prediction value of each device during the thermal processing of aerospace components. Based on the standard thermal processing process route data, the standard value of energy consumption required for processing the component is predicted as a reference value for energy consumption in actual processing;

[0076] S5: Input the real-time process data in actual hot working production into the quality prediction model to obtain the quality prediction value of the hot working component product;

[0077] S6: Based on whether the quality prediction value meets the quality requirements and the deviation between the actual equipment energy consumption value and the standard energy consumption prediction value, the particle swarm algorithm is used to adjust and optimize the process of subsequent processes to achieve coordinated control of the quality and energy consumption of hot-processed components.

[0078] An embodiment of the present invention provides a method for collaboratively controlling the thermal processing quality and energy consumption of large aerospace components, the method specifically comprising:

[0079] S1: Collect energy consumption data and process data of each device during the thermal processing of aerospace components, and combine them with existing historical energy consumption data to form an equipment energy consumption characteristic data set;

[0080] S2: Import the aerospace component model and mold model into the finite element simulation software. Based on the actual hot working process, set multiple sets of multi-pass process parameters for simulation calculation to obtain the corresponding quality data. Combine the quality inspection data of each batch of components and the corresponding process data to form a quality feature data set;

[0081] S3: Inputting the energy consumption characteristic data set and the quality characteristic data set into the adaptive fuzzy neural network to be trained respectively until the trained model converges, thereby obtaining an energy consumption prediction model for each hot processing equipment and a multi-pass quality prediction model for hot processing components;

[0082] S4: Inputting the preset thermal processing process data into the energy consumption prediction model to obtain the standard energy consumption prediction value of each device during the thermal processing of aerospace components. Based on the standard thermal processing process route data, the standard value of energy consumption required for processing the component is predicted as a reference value for energy consumption in actual processing;

[0083] S5: Input the real-time process data in actual hot working production into the quality prediction model to obtain the quality prediction value of the hot working component product;

[0084] S6: Based on whether the quality prediction value meets the quality requirements and the deviation between the actual equipment energy consumption value and the standard energy consumption prediction value, the particle swarm algorithm is used to adjust and optimize the process of subsequent processes to achieve coordinated control of the quality and energy consumption of hot-processed components.

[0085] The S1, hot working process includes blank making, die forging and heat treatment; the energy consumption data and process data of each equipment include the input power, pressing speed, pressing force, die temperature, component temperature of the forging press, the input power or gas consumption rate, heating rate, and temperature change curve in the furnace of the heating furnace or heat treatment furnace.

[0086] The S2 quality data includes the grain size, yield strength, tensile strength, etc. of the final hot-processed component; the process data includes the forging rate of multiple forging passes, component temperature before forging, die temperature, heat treatment heating rate, holding temperature, holding time, etc.

[0087] The energy consumption prediction model in S3 is developed for different hot working equipment, including open die forging presses, die forging presses, pre-forging furnaces, and heat treatment furnaces. The model input is a time series of equipment process data, and the output is a time series of equipment power and gas consumption rates. After processing, this data can be converted into the total power and gas consumption of the equipment over a period of time. This prediction model can simulate and calculate the corresponding input power or gas consumption rate when the equipment is operating according to a set of set process parameters in actual production.

[0088] Quality prediction models were developed for different output targets, including multiple prediction models that use quality indicators such as grain size, yield strength, and tensile strength as output values. The model inputs were a matrix of process parameters from the entire process. The model established a correlation between the final quality of the forging and the complete hot working process.

[0089] The method for collaboratively controlling the quality and energy consumption of large aerospace components during hot working, provided by embodiments of this invention, aims to achieve dual optimization of quality control and energy management during hot working through data-driven prediction and optimization techniques. By integrating four key steps—data collection, analysis, prediction, and optimization—this method ensures that aerospace components meet high quality standards during hot working while minimizing energy consumption and improving production efficiency and economic benefits.

[0090] First, in step S1: Energy consumption data and process data for each piece of equipment during the hot working of aerospace components are collected. The system uses sensors installed on various hot working equipment (such as blank forming machines, die forging presses, heating furnaces, and heat treatment furnaces) to monitor and record the equipment's energy consumption parameters (such as input power, pressing speed, pressing force, mold temperature, component temperature, gas consumption rate, heating rate, etc.) and process parameters in real time. This data includes not only real-time data but also combines historical energy consumption data to form a comprehensive data set of equipment energy consumption characteristics. In this way, the system can fully understand the energy consumption performance of each piece of equipment under different process conditions, providing a reliable data foundation for subsequent energy consumption prediction and optimization.

[0091] Next, in step S2: the aerospace component model and the mold model are imported into the finite element simulation software, and the system uses finite element analysis (FEA) technology to simulate the actual hot working process. By setting multiple sets of multi-pass process parameters (such as forging rate, component temperature before forging, mold temperature, heat treatment heating rate, holding temperature and holding time, etc.), the system can simulate the effects of different process parameter combinations on component quality (such as grain size, yield strength, tensile strength, etc.). At the same time, the actual quality inspection data of each batch of components is combined to form a quality feature data set. This method of combining simulation with actual data ensures the accuracy and reliability of the quality prediction model, laying a solid foundation for subsequent quality and energy consumption predictions.

[0092] In step S3: the energy consumption characteristic data set and the quality characteristic data set are respectively input into the adaptive fuzzy neural network to be trained. The system uses an advanced adaptive fuzzy neural network (AFNN) for data modeling. By training the energy consumption characteristic data set, the system can establish energy consumption prediction models for each hot working equipment. These models can accurately predict the power consumption or gas consumption rate of the equipment under different process parameters based on the time series of equipment process data. At the same time, the training of the quality characteristic data set enables the system to establish multi-pass quality prediction models for hot working components. These models can predict the key quality indicators of the components based on the complete hot working process parameters. The training process continues until the model converges to ensure the high accuracy and stability of the prediction results.

[0093] Step S4: Input the preset heat treatment process data into the energy consumption prediction model. The system uses the trained energy consumption prediction model and standard heat treatment process data to predict the standard energy consumption values ​​required for each piece of equipment during the process. These standard energy consumption values ​​serve as a reference for actual production energy consumption, helping companies make scientific decisions when setting process parameters and ensuring that energy consumption remains within a reasonable range. The system also compares these predicted values ​​with real-time energy consumption data from actual processing to promptly identify energy consumption anomalies and ensure energy-saving and efficient production.

[0094] The fuzzy neural network in S4, Figure 2 This is the structural diagram of the fuzzy neural network. The prediction model establishment process includes:

[0095] (1) Fuzzify the input parameters and establish fuzzy rules. According to the complexity and distribution characteristics of the input parameters, select the appropriate membership function and the number of membership functions. The membership function can be Gaussian membership function, triangle membership function, trapezoidal membership function, etc. The number of functions ranges from 2 to 8. The fuzzification operation can be expressed as:

[0096]

[0097] x i represents the i-th input parameter, μ ij (x i ) is the j-th membership function of the i-th input parameter, is the membership value.

[0098] (2) Fuzzy rule incentive intensity calculation: multiply any two membership degrees of any two input parameters and the output is:

[0099]

[0100] (3) The normalized calculation of the usage of each fuzzy rule ensures that the relative importance of all rules is balanced. The function is:

[0101]

[0102] (4) Calculation of the normalized regular excitation intensity linear output:

[0103]

[0104] (5) Weighted summation and defuzzification:

[0105]

[0106] (6) Fuzzy neural network parameter optimization: The parameters of the membership function in step (1) and the linear function in step (4) are adjusted by the back propagation algorithm to make the prediction model converge.

[0107] In step S5, real-time process data from actual hot working production is input into the quality prediction model. The system monitors real-time process parameters (such as forging rate, die temperature, and heating rate) and uses the trained quality prediction model to predict the quality indicators of hot-worked component products in real time. These predicted values ​​reflect the impact of current process parameters on component quality, providing a basis for timely adjustment of process parameters during production to ensure that each batch of products meets the predetermined quality standards.

[0108] Finally, in step S6, based on whether the quality prediction value meets the quality requirements and the deviation between the actual equipment energy consumption value and the standard energy consumption prediction value, the particle swarm algorithm is used to adjust and optimize the process of subsequent steps. The system uses the particle swarm optimization algorithm (PSO) to dynamically optimize the process parameters. By analyzing the deviation between the quality prediction value and the target quality value, as well as the difference between the actual energy consumption and the standard energy consumption prediction value, the particle swarm algorithm can adjust the process parameters within the framework of multi-objective optimization to achieve the coordinated optimization of quality and energy consumption. This optimization process not only improves the processing quality of the component, but also effectively reduces energy consumption, improves overall production efficiency and economic benefits, and ultimately achieves the goals of high quality and energy saving in the hot processing of large aerospace components.

[0109] Furthermore, in S6, when it is predicted that the final quality of the component is unqualified, a particle swarm algorithm is used to perform multi-objective optimization on the subsequent process. The process parameters are updated through the algorithm and input into the quality and energy consumption prediction model. The process parameters are updated with the predicted final quality of the component as a constraint and the predicted minimum required energy consumption as the goal. Finally, optimization is achieved. The objective function can be expressed as:

[0110]

[0111] Among them, X represents the multidimensional process matrix composed of the whole process parameters, Q i (X) indicates the predicted values ​​of different qualities, a i ~b i It represents the required range of different qualities, and E(X) represents the total energy consumption of the whole process, which is the sum of the energy consumption forecasts of each equipment.

[0112] In summary, the method for coordinated control of thermal processing quality and energy consumption of large aerospace components of the present invention solves key technical problems in the field of large forging processing through innovative data acquisition, analysis, prediction and optimization technologies, significantly improves the efficiency and effect of the system, and has broad application prospects and significant technological progress.

[0113] Adaptive fuzzy neural network is a hybrid intelligent algorithm that combines fuzzy logic and artificial neural network. It can use the self-learning ability of neural network and the knowledge expression ability of fuzzy logic to deal with uncertainty and ambiguity in complex systems. The basic structure is as follows Figure 2 shown.

[0114] The square nodes in the network need to learn parameters. Its structure is a five-layer feedforward network. The node function of each layer of the network structure is described as follows:

[0115] like Figure 3 As shown, an embodiment of the present invention provides a large-scale aerospace component thermal processing quality and energy consumption coordinated control system, specifically including:

[0116] Data acquisition module, which collects and processes data to build equipment energy consumption characteristic data sets and quality characteristic data sets;

[0117] Energy consumption prediction module, based on the energy consumption prediction model, obtains the standard energy consumption prediction value of each equipment in the thermal processing of aerospace components;

[0118] The quality prediction module obtains the quality prediction value of the hot-processed component product based on the quality prediction model;

[0119] The optimization module uses particle swarm optimization to perform multi-objective optimization on subsequent processes.

[0120] 1. Specific application fields or related products of the present invention.

[0121] Example based on a certain type of aluminum alloy aviation die forging

[0122] Figure 4 Hot working process of the die forging

[0123] 1. Collect the energy consumption data and process data of each device during the processing, including the input power, pressing speed, pressing force, die temperature, component temperature of the die forging hydraulic press, the input power and furnace temperature change curve of the pre-forging heating furnace and heat treatment rate, to form the energy consumption characteristic data set of each device.

[0124] 2. The forging and corresponding die models were imported into the simulation software Deform. Multiple sets of process parameters were set for each process step, including forging rates for pre-forging and final forging, initial temperatures for the forging and die, and heat treatment heating rates, holding temperatures, and holding times. Simulations were performed across multiple steps and process groups to determine the final average grain size data for the die forgings under each process group. This data was combined with the average grain size from post-production quality inspections of the die forgings and the corresponding process data to construct a characteristic data set for the average grain size of the die forgings.

[0125] 3. Establish energy consumption prediction models for forging presses, pre-forging heating furnaces, and heat treatment furnaces, and establish an average grain size prediction model for die forgings. This example details the process of establishing an energy consumption prediction model for pre-forging heating furnaces:

[0126] Since the heating furnace only works during heating and insulation, only the data of heating and insulation stages are used for model training. Fuzzy neural network is used for modeling, and the membership function is Gaussian function, which is expressed as

[0127]

[0128] The input variables of the model are set temperature, average temperature in the furnace, and heating rate, and the output variable is power. The number of membership functions for each input variable is set to 3. The input data is processed through the fuzzy layer, rule layer, normalization layer, and linear layer to obtain the output result. The model is trained through the hybrid optimization algorithm, and the energy consumption prediction model E for the pre-forging heating furnace is finally obtained. 加热 (X 加热 ).

[0129] According to the above method, the energy consumption prediction model E of the die forging press is obtained. 压机 (X 压机 ), heat treatment furnace energy consumption prediction model E 热处理 (X 热处理 ) and the average grain size prediction model D(X) for die forgings.

[0130] 4. Input the standard process data of the aluminum alloy die forgings into the energy consumption prediction models of the die forging press, pre-forging heating furnace, and heat treatment furnace respectively, and calculate the standard energy consumption value of each equipment in each process as a reference value for actual processing.

[0131] 5. In actual processing, the actual process data of the completed process and the standard process data of the subsequent process are combined and input into the average grain size prediction model of the die forging to obtain the final average grain size prediction value of the current die forging.

[0132] 6. Based on the deviation between the predicted and required average grain size, and the deviation between the actual total energy consumption of the equipment and the predicted standard total energy consumption, the particle swarm algorithm is used to adjust and optimize the process of subsequent steps to achieve coordinated control of the quality and energy consumption of hot-processed components.

[0133] Optimization is mainly required for the following two situations:

[0134] (1) The predicted average grain size does not meet the requirements. In this example, the grain size requirement for the aviation aluminum alloy die forging is 7-8, that is, the average grain size requirement range is 0.016-0.032mm. When the predicted average grain size exceeds the required range, subsequent process optimization is required.

[0135] (2) The predicted average grain size meets the requirement, but the actual energy consumption of the previous process exceeds the standard energy consumption value by a certain range. In this example, the actual energy consumption of each process is allowed to exceed the standard energy consumption by 10%. If the actual energy consumption is too high, subsequent process optimization is required.

[0136] The particle swarm algorithm is used for optimization, the purpose of which is to find a process combination solution that meets the average grain size requirements and has the lowest energy consumption. The optimization objective function is:

[0137]

[0138] The input X of D(X) is the process data of the completed process and the process data of the subsequent processes, and the optimization object is the process combination of the subsequent processes. E(X) represents the total energy consumption prediction value, which is the sum of the actual energy consumption of the completed process and the predicted energy consumption of the subsequent processes.

[0139] After a pre-forging, the quality prediction module predicts that the final average grain size of the aluminum alloy forging will be greater than 0.032 mm. The control system starts the optimization module to optimize the subsequent final forging and heat treatment processes. Figure 5 Iterative optimization prediction results for the average grain size, Figure 6 Iterative optimization prediction results for total energy consumption.

[0140] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware portion can be implemented using dedicated logic; the software portion can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those skilled in the art will appreciate that the above-mentioned devices and methods can be implemented using computer-executable instructions and / or contained in processor control code, for example, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, or can be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

[0141] The above description is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with this technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered by the scope of protection of the present invention.

Claims

1. A method for collaboratively controlling the hot working quality and energy consumption of large aerospace forgings, characterized in that: The method specifically includes: S1: Collect energy consumption data and process data of each device during the hot working of aerospace forgings, and combine them with existing historical energy consumption data to form an equipment energy consumption characteristic data set; S2: Import the aerospace forging model and die model into the finite element simulation software. Based on the actual hot working process, set multiple sets of multi-pass process parameters for simulation calculation to obtain the corresponding quality data. Combine the quality inspection data of each batch of forgings and the corresponding process data to form a quality feature data set; S3: Inputting the energy consumption characteristic data set and the quality characteristic data set into the adaptive fuzzy neural network to be trained respectively until the trained model converges, thereby obtaining an energy consumption prediction model for each hot working equipment and a multi-pass quality prediction model for hot working forgings; S4: Inputting the preset hot working process data into the energy consumption prediction model to obtain the standard energy consumption prediction value of each equipment in the hot working process of aerospace forgings. Based on the standard hot working process route data, the standard value of energy consumption required for processing the forgings is predicted as a reference value for energy consumption in actual processing; S5: Input the real-time process data in actual hot working production into the quality prediction model to obtain the quality prediction value of the hot working forging product; S6: Based on whether the quality prediction value meets the quality requirements and the deviation between the actual equipment energy consumption value and the standard energy consumption prediction value, the particle swarm algorithm is used to adjust and optimize the process of subsequent processes to achieve coordinated control of the quality and energy consumption of hot-processed forgings; The heat treatment process in step S1 includes blank making, die forging, and heat treatment; the energy consumption data and process data of each device include the input power, pressing speed, pressing force, die temperature, forging temperature of the forging press, the input power or gas consumption rate, heating rate, and temperature change curve of the heating furnace or heat treatment furnace; The S2, quality data includes the grain size, yield strength, and tensile strength of the final product of the hot-processed forging; the process data includes the forging rate of multiple forging passes, the forging temperature before forging, the die temperature, the heat treatment heating rate, the holding temperature, and the holding time; The energy consumption prediction model S3 is established for different hot working equipment, including: open die forging presses, die forging presses, pre-forging heating furnaces, and heat treatment furnaces. The model input is the time series of equipment process data, and the output is the time series of equipment power / gas consumption rate. After processing, it can be converted into the total power consumption / gas consumption of the equipment over a period of time. This prediction model can simulate and calculate the corresponding input power or gas consumption rate when the equipment is operated according to a series of set process parameters in actual production; Quality prediction models are established for different output targets, including multiple prediction models with grain size, yield strength, and tensile strength quality indicators as output values. The model input is the full-process process parameter matrix. The model establishes the relationship between the final quality of the forging and the complete hot working process technology.

2. The method for coordinated control of hot working quality and energy consumption of large aerospace forgings according to claim 1, characterized in that: In S6, when it is predicted that the final quality of the forging is unqualified, the particle swarm algorithm is used to perform multi-objective optimization on the subsequent process, and the process parameters are updated through the algorithm. The process parameters are input into the quality and energy consumption prediction model. The process parameters are updated with the predicted final quality of the forging as a constraint and the predicted minimum energy consumption as the goal. Finally, optimization is achieved. The objective function can be expressed as: Among them, X represents the multidimensional process matrix composed of the whole process parameters, Q i (X) indicates the predicted values ​​of different qualities, a i ~b i It represents the required range of different qualities, and E(X) represents the total energy consumption of the whole process, which is the sum of the energy consumption forecasts of each equipment.

3. A system based on the method for coordinated control of hot working quality and energy consumption of large aerospace forgings according to any one of claims 1 to 2, characterized in that: The system specifically includes: a data acquisition unit configured to collect energy consumption data and process data of various equipment during the hot working process of aerospace forgings, and to combine the existing historical energy consumption data to form an equipment energy consumption characteristic data set; A finite element simulation module is configured to receive aerospace forging models and die models, import them into finite element simulation software, set multiple sets of multi-pass process parameters based on the actual hot working process, perform simulation calculations, generate corresponding quality data, and combine the quality inspection data and process data of each batch of forgings to form a quality feature data set; An adaptive fuzzy neural network training module is configured to input the equipment energy consumption characteristic data set and the quality characteristic data set into the adaptive fuzzy neural network to be trained, respectively, for training until the model converges, thereby generating an energy consumption prediction model for each hot working equipment and a multi-pass quality prediction model for hot working forgings; an energy consumption prediction module configured to input preset thermal processing process data into the energy consumption prediction model to generate a standard energy consumption prediction value for each device in the thermal processing process of aerospace forgings; a quality prediction module configured to input real-time process data in actual hot working production into the quality prediction model to generate a quality prediction value of the hot working forging product; The optimization control unit is configured to use the particle swarm algorithm to adjust and optimize the process parameters of subsequent processes based on whether the quality prediction value meets the quality requirements and the deviation between the actual equipment energy consumption value and the standard energy consumption prediction value, so as to achieve coordinated control of the quality and energy consumption of hot-processed forgings.

4. The system according to claim 3, wherein: The data acquisition unit further comprises: Energy consumption data acquisition module, used to collect energy consumption data and process data of various equipment such as input power, pressing speed, pressing force, die temperature, forging temperature of forging press, input power or gas consumption rate, heating rate, and temperature change curve of heating furnace or heat treatment furnace; The process data acquisition module is used to collect process parameter data of multi-pass forging, such as forging rate, forging temperature before forging, die temperature, heat treatment heating rate, holding temperature and holding time.

5. The system according to claim 3, wherein: The optimization control unit further comprises: Deviation analysis module, used to calculate the deviation between the quality prediction value and the quality requirement value, as well as the deviation between the actual equipment energy consumption value and the standard energy consumption prediction value; The process parameter adjustment module dynamically adjusts the forging rate, pressing speed, and heating rate process parameters of subsequent processes based on the deviation analysis results using a particle swarm algorithm to optimize the quality and energy consumption during the hot working process.

Images