Intelligent optimization method and system for irradiation crosslinking process of flexible armor wire

Abstract

The invention relates to an intelligent optimization method and system for a flexible armor wire irradiation crosslinking process, and the method comprises the steps: S1, building a sensing network of the wire irradiation crosslinking process through an industrial internet, and collecting the real-time data of wire irradiation crosslinking; s2, constructing a mathematical relationship model between the process parameters and the material performance through a physical information neural network; s3, in combination with a finite element simulation method, simulating effects of a thermal field and a dose field in an irradiation process, and correcting unreasonable hypotheses in the mathematical relationship model; s4, on the basis of the corrected mathematical relationship model, under the multi-target process requirements, balancing different targets through an optimization algorithm and based on data acquired by a data real-time acquisition system, and determining optimal process parameters; and S5, after the optimal process parameters are applied to the production line, an optimization result is obtained through a data real-time acquisition system, and whether a production target is met is judged, so that intelligent closed loop is completed. The production efficiency is effectively improved, and the cost and risk of quality management are reduced.

Description

Technical Field

[0001] The present invention relates to the field of intelligent production control, and in particular to an intelligent optimization method and system for a radiation cross-linking process of flexible armor wires. Background Art

[0002] In modern industrial manufacturing, flexible armored wires are widely used in power and communication transmission in various complex environments, and are valued for their excellent mechanical and electrical properties. In order to further improve the durability and reliability of such wires, the radiation cross-linking process has become a key technical means. Through electron beam irradiation, the polymer materials in the wires can undergo cross-linking reactions, thereby enhancing their heat resistance, mechanical strength and chemical stability. However, the optimization of the radiation cross-linking process faces huge challenges, mainly due to the sensitivity and complexity of the process parameters to the material properties. The traditional radiation cross-linking process often relies on experience and repeated experiments, and lacks systematic and intelligent optimization methods. Summary of the invention

[0003] In order to solve the above problems, the purpose of the present invention is to provide an intelligent optimization method and system for the radiation cross-linking process of flexible armor wires, which realizes the high automation and intelligence of the production process of the radiation cross-linking process of flexible armor wires, effectively improves production efficiency, and reduces the cost and risk of quality management.

[0004] To achieve the above object, the present invention adopts the following technical solutions: An intelligent optimization method for a flexible armor wire radiation cross-linking process comprises the following steps: S1: Establish a sensing network for wire irradiation cross-linking process through the industrial Internet to collect real-time data of wire irradiation cross-linking; S2: Construct a mathematical relationship model between process parameters and material properties through physical information neural networks PINNs; S3: Combine finite element simulation method to simulate the effects of thermal field and dose field during irradiation and correct unreasonable assumptions in mathematical relationship model; S4: Based on the modified mathematical relationship model, under the multi-objective process requirements, based on the data collected by the real-time data acquisition system, the optimization algorithm is used to balance different objectives and determine the optimal process parameters; S5: After applying the optimal process parameters to the production line, the optimization results are obtained through the real-time data acquisition system, and it is determined whether they meet the production goals, thus completing the intelligent closed loop.

[0005] Furthermore, S1 is specifically: Install a radiometer, electron beam monitoring sensor, speed and vibration sensor inside the irradiation equipment to collect irradiation dose D, electron beam energy E, beam intensity I, conveying speed v and equipment vibration status; An operating parameter monitoring device and a profile measuring instrument are set at the conveying channel to collect material diameter d, thickness δ and thermophysical parameters; the thermophysical parameters include material density ρ, specific heat capacity C p , thermal conductivity k; Temperature and humidity sensors are arranged outside the production line to collect the ambient temperature T env and ambient humidity H env ; and monitoring cross-linking hot spots with infrared thermal imaging equipment; The devices and sensors are connected via industrial Ethernet, and MQTT is used to achieve low-latency and high-reliability data transmission, and the collected data is transmitted to the monitoring database.

[0006] Furthermore, the physical information neural network PINNs include an input layer, a hidden layer, physical constraints, a loss function, and an output layer; In the radiation cross-linking process, the relationship between various physical parameters and material properties is described by the following equation: ; ; ; Where, t is time; is the radiation absorption coefficient; is the irradiation temperature rise; is the cross-link density; Material performance indicators include flexibility F r And fire resistance T r , and flexibility F r and fire resistance T r The relationship with crosslink density is obtained by fitting: ; ; in, a 1 、a 2 、b 1 、b 2 is the fitting parameter; Physical information neural network PINNs takes the relationship between each physical parameter and material performance as the physical constraint of the neural network, and combines real experimental data to optimize the neural network. The input layer inputs the process parameter X: X=[I,E,v,T env ,ρ,C p ]; The hidden layer fits the nonlinear relationship through the fully connected feedforward neural network hidden layer NN(·): ; Among them, θ is the physical information neural network PINNs parameter; Physical constraints, including radiation dose constraints Thermal field constraints , the relationship between crosslink density and performance, specifically: ; ; ; ; Among them, L crosslink is the constraint of cross-linking density and dosage; L performance is the material performance constraint; L fire For fire resistance constraints; is the square of the second norm of the deviation; PINNs loss function is a joint loss function that combines physical constraints and experimental data:

[0007] Among them, w1, w2, w3, w4, and w5 are weight coefficients that adjust the impact of different loss terms; L-BFGS is used to optimize θ and minimize the PINNs loss function to obtain the trained PINNs model.

[0008] Furthermore, L-BFGS is used to optimize θ, as follows: The random initialization parameter is θ0, and the initial value of the H matrix is ​​set to the unit matrix; Forward propagation, calculate the joint loss function L of the current parameter θt total (θ t ); Back propagation, calculate the objective function gradient ; Hessian update, using L-BFGS to dynamically construct the current H matrix H t The approximate value of according to Iterate; H t Update approximately as follows: ; ; ; ; Among them, ρ t is the scaling factor; s t is the parameter change; y t is the gradient change; T is the transpose; If the joint loss function meets the convergence conditions, output the final parameter θ; otherwise, repeat the above steps.

[0009] Furthermore, S3 is specifically: S31: Establish a physical model to describe the mathematical relationship between the thermal field and the dose field during irradiation, including the dose field distribution model and the thermal field distribution model; During the irradiation process, the energy deposition of the electron beam in the material follows the Bethe-Bloch equation, and the dose field distribution model is described as: ; Where E is the electron beam energy; S is the material depth; is a constant related to the electron beam energy; Z and A are the atomic number and mass number of the material respectively; ρ is the material density; is the electron speed; Dose distribution was simulated by Monte Carlo method: ; in, f(x,y,z) is the normalized spatial distribution function; (x,y,z) is the spatial coordinate, is the dose field; The thermal field distribution model is described by the heat conduction equation: ; in, (x, y, z, t) is the temperature distribution; Q(x, y, z, t) is the heat source term, which is determined by the dose field Converted into heat; The relationship between the heat source term Q and the dose field is: ; Where η is the energy conversion efficiency, which indicates the ratio of absorbed dose converted into heat; S32: Based on the FEM numerical method, solve the thermal field and dose field distribution in the material and obtain the dynamic change results in space and time; S33: Compare the FEM simulation results with the theoretical predictions of the mathematical relationship model, analyze the differences, correct unreasonable physical assumptions, optimize the mathematical relationship model, and the corrected mathematical relationship model is used to guide the physical constraints of the PINNs model and the experimental data fitting.

[0010] Furthermore, based on the FEM numerical method, the thermal field and dose field distribution in the material are solved, and the dynamic change results in space and time are obtained, as follows: Establish a three-dimensional geometric model based on the actual shape and size of the material; Discretize the geometric model into a finite element mesh and control the cell size to balance accuracy and computational efficiency; Set the boundary conditions and initial conditions: Initial conditions: initial temperature T'(x,y,z,t=0)=T env ; Boundary conditions: Convection boundary conditions: ; Where h is the convective heat transfer coefficient; FEM software is used to solve the dynamic distribution of dose field and thermal field to obtain D'(x,y,z,t) and T'(x,y,z,t).

[0011] Further, S33 is as follows: The dose field D calculated by FEM FEM (x, y, z, t) and the dose field D predicted by the mathematical relationship model model (x,y,z,t) is compared to calculate the dose field difference ΔD(x,y,z,t): ; The thermal field T calculated by FEM FEM (x, y, z, t) and the thermal field T predicted by the mathematical relationship model model (x,y,z,t) is compared to calculate the thermal field difference ΔT(x,y,z,t): ; If the thermal field distribution error is greater than the threshold, the η assumption is unreasonable; by fitting the heat source term Q of FEM FEM (x,y,z,t) and dose field D FEM (x,y,z,t), recalculate η: ; If the temperature field error on the material surface is greater than the threshold, the assumptions of the convection heat transfer coefficient h or the ambient temperature Tenv are unreasonable and need to be reset; If the dose field error is greater than the threshold, the spatial distribution function f(x, y, z) of the Monte Carlo simulation needs to be re-evaluated; Corrected dose field distribution D opt ( x , y ,z , t ): ; Among them, f opt (x, y, z) is the spatial distribution function of the dose corrected by FEM simulation; Corrected heat conduction equation: ; in, η opt is the energy conversion efficiency after correction by FEM simulation, Represents the gradient operator.

[0012] Furthermore, S4 is specifically: The objective function and constraint conditions are constructed based on the modified mathematical relationship model; the objective function F is: ; ; ; ; in, They are respectively for dose uniformity optimization, temperature rise minimization and production efficiency maximization; is the temperature rise safety threshold; is the maximum temperature rise; , and is the weight coefficient; The constraints include dose range constraints, material temperature constraints and process parameter range constraints, where: Dose range constraint: Ensure that the dose within the material does not exceed the set range: Material temperature constraint: Material temperature must be limited below the safe value; Process parameter range constraints: beam intensity I, electron beam energy E, and transport speed v meet the actual equipment capabilities; Collect data feedback in real time to provide a basis for process adjustment; The multi-objective optimization algorithm NSGA-II is used to find the optimal balance between multi-objective conflicts; the Pareto optimal solution of multi-objective optimization is obtained to obtain the optimal combination of process parameters.

[0013] Furthermore, S5 is specifically: Apply the optimal process parameters obtained through the optimization algorithm to the production line; Install a real-time data acquisition system on the production line to capture performance test data of processed materials, including cross-linking density, flexibility, and fire resistance; Compare the test data with the production target value to determine whether the test indicators meet the process requirements; If the inspection data deviates from the target, the process parameters are adjusted through the genetic algorithm and applied to the production line again to form a closed loop until the production target is achieved.

[0014] An intelligent optimization system for a flexible armor wire radiation cross-linking process comprises a processor, a memory and a computer program stored in the memory. When the processor executes the computer program, it specifically performs the steps in the intelligent optimization method for a flexible armor wire radiation cross-linking process as described above.

[0015] The present invention has the following beneficial effects: 1. The present invention realizes the high automation and intelligence of the production process of the radiation cross-linking process of the flexible armor wire, effectively improves the production efficiency, and reduces the cost and risk of quality management; 2. The present invention uses FEM simulation to simulate the thermal field and dose field in the irradiation process in detail. Combined with the theoretical prediction of the mathematical relationship model, it can discover and correct the unreasonable physical assumptions in the model, significantly improve the physical consistency and prediction accuracy of the model, and provide a reliable basis for the optimization and control of the irradiation process; 3. The present invention integrates the modified mathematical relationship model through a real-time acquisition system and a multi-objective optimization method, dynamically optimizes the irradiation process parameters, and obtains the Pareto optimal solution set for users to choose on the premise of meeting conflicting requirements such as dose uniformity, temperature rise safety, and production efficiency, thereby achieving a dynamic balance between efficiency and safety. Based on the optimal process parameters obtained through optimization, after applying it to the production line, dynamic judgment is made through real-time detection data, the optimization results are fed back, and it is judged whether it meets the production goals. It is effectively used in the irradiation processing line to improve the overall production efficiency and quality control level. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0017] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments: refer to Figure 1 In this embodiment, an intelligent optimization method for the radiation cross-linking process of flexible armor wire is provided, comprising the following steps: S1: Establish a sensing network for wire irradiation cross-linking process through the industrial Internet to collect real-time data of wire irradiation cross-linking; S2: Construct a mathematical relationship model between process parameters and material properties through physical information neural networks PINNs; S3: Combine finite element simulation method to simulate the effects of thermal field and dose field during irradiation and correct unreasonable assumptions in mathematical relationship model; S4: Based on the modified mathematical relationship model, under the multi-objective process requirements, based on the data collected by the real-time data acquisition system, the optimization algorithm is used to balance different objectives and determine the optimal process parameters; S5: After applying the optimal process parameters to the production line, the optimization results are obtained through the real-time data acquisition system, and it is determined whether they meet the production goals, thus completing the intelligent closed loop.

[0018] In this embodiment, S1 is specifically: Install a radiometer, electron beam monitoring sensor, speed and vibration sensor inside the irradiation equipment to collect irradiation dose D, electron beam energy E, beam intensity I, conveying speed v and equipment vibration status; An operating parameter monitoring device and a profile measuring instrument are set at the conveying channel to collect material diameter d, thickness δ and thermophysical parameters; the thermophysical parameters include material density ρ, specific heat capacity C p , thermal conductivity k; Temperature and humidity sensors are arranged outside the production line to collect the ambient temperature T env and ambient humidity H env ; and monitoring cross-linking hot spots with infrared thermal imaging equipment; The devices and sensors are connected via industrial Ethernet, and MQTT is used to achieve low-latency and high-reliability data transmission, and the collected data is transmitted to the monitoring database.

[0019] In this embodiment, the physical information neural network PINNs includes an input layer, a hidden layer, a physical constraint, a loss function and an output layer; In the radiation cross-linking process, the relationship between various physical parameters and material properties is described by the following equation: ; ; ; Where, t is time; is the radiation absorption coefficient; is the irradiation temperature rise; is the cross-link density; Material performance indicators such as flexibility F r and fire resistance T r Crosslinking density The relationship is obtained by fitting: ; ; in,a 1 、a 2 、b 1 、b 2 is the fitting parameter; Physical information neural network PINNs takes the relationship between each physical parameter and material performance as the physical constraint of the neural network, and combines real experimental data to optimize the neural network. The input layer inputs the process parameter X: X=[I,E,v,T env ,ρ,C p ]; The hidden layer fits the nonlinear relationship through the fully connected feedforward neural network hidden layer NN(·): ; Among them, θ is the parameter of physical information neural network PINNs; Physical constraints, including radiation dose constraints Thermal field constraints , the relationship between crosslink density and performance, specifically: ; ; ; ; Among them, L crosslink is the constraint of cross-linking density and dosage; L performance is the material performance constraint; L fire For fire resistance constraints; is the square of the second norm of the deviation; PINNs loss function is a joint loss function that combines physical constraints and experimental data:

[0020] Among them, w1, w2, w3, w4, and w5 are weight coefficients that adjust the impact of different loss terms.

[0021] L-BFGS is used to optimize θ and minimize the PINNs loss function to obtain the trained PINNs model.

[0022] In this embodiment, L-BFGS is used to optimize θ, as follows: The random initialization parameter is θ0, and the initial value of the H matrix is ​​set to the unit matrix; Forward propagation, calculate the joint loss function L of the current parameter θt total (θ t ); Back propagation, calculate the objective function gradient ; Hessian update, using L-BFGS to dynamically construct the current H matrix H t The approximate value of according to Iterate; H t Update approximately as follows: ; ; ; ; Among them, ρ t is the scaling factor; s t is the parameter change; y t is the gradient change; T is the transpose; If the joint loss function meets the convergence conditions, output the final parameter θ; otherwise, repeat the above steps.

[0023] In this embodiment, S3 is specifically: S31: Establish a physical model to describe the mathematical relationship between the thermal field and the dose field during irradiation, including the dose field distribution model and the thermal field distribution model; During the irradiation process, the energy deposition of the electron beam in the material follows the Bethe-Bloch equation , describing the dose field distribution model as: ; Where E is the electron beam energy; S is the material depth; is a constant related to the electron beam energy; Z and A are the atomic number and mass number of the material respectively; ρ is the material density; is the electron speed; Dose distribution was simulated by Monte Carlo method: ; in, f(x,y,z) is the normalized spatial distribution function; (x,y,z) is the spatial coordinate, is the dose field; The thermal field distribution model is described by the heat conduction equation: ; in, (x, y, z, t) is the temperature distribution; Q(x, y, z, t) is the heat source term, which is determined by the dose field Converted into heat; The relationship between the heat source term Q and the dose field is: ; Where η is the energy conversion efficiency, which indicates the ratio of absorbed dose converted into heat; S32: Based on the FEM numerical method, solve the thermal field and dose field distribution in the material and obtain the dynamic change results in space and time; S33: Compare the FEM simulation results with the theoretical predictions of the mathematical relationship model, analyze the differences, correct unreasonable physical assumptions, optimize the mathematical relationship model, and the corrected mathematical relationship model is used to guide the physical constraints of the PINNs model and the experimental data fitting.

[0024] In this embodiment, based on the FEM numerical method, the thermal field and dose field distribution in the material are solved to obtain the dynamic change results in space and time, as follows: Establish a three-dimensional geometric model based on the actual shape and size of the material; Discretize the geometric model into a finite element mesh and control the cell size to balance accuracy and computational efficiency; Set the boundary conditions and initial conditions: Initial conditions: initial temperature T'(x,y,z,t=0)=T env ; Boundary conditions: Convection boundary conditions (such as heat dissipation on the surface of the material): ; Where h is the convective heat transfer coefficient; Use FEM software (such as COMSOL, ANSYS or ABAQUS) to solve the dynamic distribution of dose field and thermal field and obtain D'(x,y,z,t) and T'(x,y,z,t).

[0025] In this embodiment, S33 is specifically as follows: The dose field D calculated by FEM FEM (x, y, z, t) and the dose field D predicted by the mathematical relationship model model (x,y,z,t) is compared to calculate the dose field difference ΔD(x,y,z,t): ; The thermal field T calculated by FEM FEM (x, y, z, t) and the thermal field T predicted by the mathematical relationship model model (x,y,z,t) is compared to calculate the thermal field difference ΔT(x,y,z,t): ; If the thermal field distribution error is greater than the threshold, the η assumption is unreasonable. By fitting the heat source term Q FEM (x,y,z,t) and dose field D FEM (x,y,z,t), recalculate η: ; If the temperature field error on the material surface is greater than the threshold, the assumptions of the convection heat transfer coefficient h or the ambient temperature Tenv are unreasonable and need to be reset; If the dose field error is greater than the threshold, the spatial distribution function f(x, y, z) of the Monte Carlo simulation needs to be re-evaluated; Corrected dose field distribution D opt ( x , y , z , t ): ; Among them, f opt (x, y, z) is the spatial distribution function of the dose corrected by FEM simulation; Corrected heat conduction equation: ; in, η opt is the energy conversion efficiency after correction by FEM simulation, Represents the gradient operator.

[0026] In this embodiment, S4 is specifically: The objective function and constraint conditions are constructed based on the modified mathematical relationship model; the objective function F is: ; ; ; ; in, They are respectively for dose uniformity optimization, temperature rise minimization and production efficiency maximization; is the temperature rise safety threshold; is the maximum temperature rise; , and is the weight coefficient; The constraints include dose range constraints, material temperature constraints and process parameter range constraints, where: Dose range constraint: Ensure that the dose within the material does not exceed the set range: Material temperature constraint: Material temperature must be limited below the safe value; Process parameter range constraints: beam intensity I, electron beam energy E, and transport speed v meet the actual equipment capabilities; Collect data feedback in real time to provide a basis for process adjustment; The multi-objective optimization algorithm NSGA-II is used to find and balance the multi-objective conflicts; the Pareto optimal solution of multi-objective optimization is obtained (a set of numerical solutions that balance different objectives for operation selection) to obtain the optimal combination of process parameters (such as beam intensity, electron beam energy, and transport speed).

[0027] In this embodiment, S5 is specifically: Apply the optimal process parameters (beam intensity, electron beam energy, conveying speed) obtained by the optimization algorithm to the production line; Install a real-time data acquisition system on the production line to capture performance test data of processed materials, including cross-linking density, flexibility, and fire resistance; Compare the test data with the production target value to determine whether the test indicators meet the process requirements; If the inspection data deviates from the target, the process parameters are adjusted through the genetic algorithm and applied to the production line again to form a closed loop until the production target is achieved.

[0028] An intelligent optimization system for a flexible armor wire radiation cross-linking process comprises a processor, a memory and a computer program stored in the memory. When the processor executes the computer program, the steps in the intelligent optimization method for a flexible armor wire radiation cross-linking process as described above are specifically performed.

[0029] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0030] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0031] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0032] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0033] The above is only a preferred embodiment of the present invention, and does not limit the present invention in other forms. Any technician familiar with the profession may use the above disclosed technical content to change or modify it into an equivalent embodiment with equivalent changes. However, any simple modification, equivalent change and modification made to the above embodiment according to the technical essence of the present invention without departing from the technical solution of the present invention still belongs to the protection scope of the technical solution of the present invention.

Claims

1. An intelligent optimization method for the radiation cross-linking process of flexible armor wires, characterized in that: The following steps are involved: S1: Establish a sensing network for wire irradiation cross-linking process through the industrial Internet to collect real-time data of wire irradiation cross-linking; S2: Construct a mathematical relationship model between process parameters and material properties through physical information neural networks PINNs; S3: Combine finite element simulation method to simulate the effects of thermal field and dose field during irradiation and correct unreasonable assumptions in mathematical relationship model; S4: Based on the modified mathematical relationship model, under the multi-objective process requirements, based on the data collected by the real-time data acquisition system, the optimization algorithm is used to balance different objectives and determine the optimal process parameters; S5: After applying the optimal process parameters to the production line, the optimization results are obtained through the real-time data acquisition system, and it is determined whether they meet the production goals, thus completing the intelligent closed loop.

2. According to claim 1, the intelligent optimization method for the radiation cross-linking process of flexible armor wire is characterized in that: The S1 is specifically: Install a radiometer, electron beam monitoring sensor, speed and vibration sensor inside the irradiation equipment to collect irradiation dose D, electron beam energy E, beam intensity I, conveying speed v and equipment vibration status; An operating parameter monitoring device and a profile measuring instrument are set at the conveying channel to collect material diameter d, thickness δ and thermophysical parameters; the thermophysical parameters include material density ρ, specific heat capacity C p , thermal conductivity k; Temperature and humidity sensors are arranged outside the production line to collect the ambient temperature T env and ambient humidity H env ; and monitoring cross-linking hot spots with infrared thermal imaging equipment; The devices and sensors are connected via industrial Ethernet, and MQTT is used to achieve low-latency and high-reliability data transmission, and the collected data is transmitted to the monitoring database.

3. The intelligent optimization method for the radiation cross-linking process of flexible armor wire according to claim 2 is characterized in that: The physical information neural network PINNs includes an input layer, a hidden layer, a physical constraint, a loss function and an output layer; In the radiation cross-linking process, the relationship between various physical parameters and material properties is described by the following equation: ； ； ； Where, t is time; is the radiation absorption coefficient; is the irradiation temperature rise; is the cross-link density; Material performance indicators include flexibility F r and fire resistance T r , and flexibility F r and fire resistance T r The relationship with crosslink density is obtained by fitting: ； ； in, a 1 、a 2 、b 1 、b 2 is the fitting parameter; Physical information neural network PINNs takes the relationship between each physical parameter and material performance as the physical constraint of the neural network, and combines real experimental data to optimize the neural network. The input layer inputs the process parameter X: X=[I,E,v,T env ,ρ,C p ]; The hidden layer fits the nonlinear relationship through the fully connected feedforward neural network hidden layer NN(·): ； Among them, θ is the physical information neural network PINNs parameter; Physical constraints, including radiation dose constraints , Thermal field constraints , the relationship between crosslink density and performance, specifically: ； ； ； ； Among them, L crosslink is the constraint of cross-linking density and dosage; L performance is the material performance constraint; L fire For fire resistance constraints; is the square of the second norm of the deviation; PINNs loss function is a joint loss function that combines physical constraints and experimental data: Among them, w1, w2, w3, w4, and w5 are weight coefficients that adjust the impact of different loss terms; L-BFGS is used to optimize θ and minimize the PINNs loss function to obtain the trained PINNs model.

4. The intelligent optimization method for the radiation cross-linking process of flexible armor wire according to claim 3 is characterized in that: The L-BFGS optimization of θ is as follows: The random initialization parameter is θ0, and the initial value of the H matrix is ​​set to the unit matrix; Forward propagation, calculate the joint loss function L of the current parameter θt total (θ t ); Back propagation, calculate the objective function gradient ; Hessian update, using L-BFGS to dynamically construct the current H matrix H t The approximate value of according to Iterate; H t Update approximately as follows: ； ； ; ; Among them, ρ t is the scaling factor; s t is the parameter change; y t is the gradient change; T is the transpose; If the joint loss function meets the convergence conditions, output the final parameter θ; otherwise, repeat the above steps.

5. The intelligent optimization method for the radiation cross-linking process of flexible armor wire according to claim 4 is characterized in that: The S3 is specifically: S31: Establish a physical model to describe the mathematical relationship between the thermal field and the dose field during irradiation, including the dose field distribution model and the thermal field distribution model; During the irradiation process, the energy deposition of the electron beam in the material follows the Bethe-Bloch equation , describing the dose field distribution model as: ； Where E is the electron beam energy; S is the material depth; is a constant related to the electron beam energy; Z and A are the atomic number and mass number of the material respectively; ρ is the material density; is the electron speed; Dose distribution was simulated by Monte Carlo method: ； in, f(x,y,z) is the normalized spatial distribution function; (x,y,z) is the spatial coordinate, is the dose field; The thermal field distribution model is described by the heat conduction equation: ； in, (x, y, z, t) is the temperature distribution; Q(x, y, z, t) is the heat source term, which is determined by the dose field Converted into heat; The relationship between the heat source term Q and the dose field is: ； Where η is the energy conversion efficiency, which indicates the ratio of absorbed dose converted into heat; S32: Based on the FEM numerical method, solve the thermal field and dose field distribution in the material and obtain the dynamic change results in space and time; S33: Compare the FEM simulation results with the theoretical predictions of the mathematical relationship model, analyze the differences, correct unreasonable physical assumptions, optimize the mathematical relationship model, and the corrected mathematical relationship model is used to guide the physical constraints of the PINNs model and the experimental data fitting.

6. The intelligent optimization method for the radiation cross-linking process of flexible armor wire according to claim 5 is characterized in that: Based on the FEM numerical method, the thermal field and dose field distribution in the material are solved to obtain the dynamic change results in space and time, as follows: Establish a three-dimensional geometric model based on the actual shape and size of the material; Discretize the geometric model into a finite element mesh and control the cell size to balance accuracy and computational efficiency; Set the boundary conditions and initial conditions: Initial conditions: initial temperature T'(x,y,z,t=0)=T env ; Boundary conditions: Convection boundary conditions: ； Where h is the convective heat transfer coefficient; FEM software is used to solve the dynamic distribution of dose field and thermal field to obtain D'(x,y,z,t) and T'(x,y,z,t).

7. The intelligent optimization method for the radiation cross-linking process of flexible armor wire according to claim 5 is characterized in that: The S33 is specifically as follows: The dose field D calculated by FEM FEM (x, y, z, t) and the dose field D predicted by the mathematical relationship model model (x,y,z,t) is compared to calculate the dose field difference ΔD(x,y,z,t): ； The thermal field T calculated by FEM FEM (x, y, z, t) and the thermal field T predicted by the mathematical relationship model model (x,y,z,t) is compared to calculate the thermal field difference ΔT(x,y,z,t): ； If the thermal field distribution error is greater than the threshold, the η assumption is unreasonable; by fitting the heat source term Q of FEM FEM (x,y,z,t) and dose field D FEM (x,y,z,t), recalculate η: ； If the temperature field error on the material surface is greater than the threshold, the assumptions of the convection heat transfer coefficient h or the ambient temperature Tenv are unreasonable and need to be reset; If the dose field error is greater than the threshold, the spatial distribution function f(x, y, z) of the Monte Carlo simulation needs to be re-evaluated; Corrected dose field distribution D opt ( x , y , z , t ): ; Among them, f opt (x, y, z) is the spatial distribution function of the dose corrected by FEM simulation; Corrected heat conduction equation: ; in, η opt is the energy conversion efficiency after correction by FEM simulation, Represents the gradient operator.

8. The intelligent optimization method for the radiation cross-linking process of flexible armor wire according to claim 7 is characterized in that: The S4 is specifically: The objective function and constraint conditions are constructed based on the modified mathematical relationship model; the objective function F is: ； ； ； ； in, They are respectively for dose uniformity optimization, temperature rise minimization and production efficiency maximization; is the temperature rise safety threshold; is the maximum temperature rise; , and is the weight coefficient; The constraints include dose range constraints, material temperature constraints and process parameter range constraints, where: Dose range constraint: Ensure that the dose within the material does not exceed the set range: Material temperature constraint: Material temperature must be limited below the safe value; Process parameter range constraints: beam intensity I, electron beam energy E, and transport speed v meet the actual equipment capabilities; Collect data feedback in real time to provide a basis for process adjustment; The multi-objective optimization algorithm NSGA-II is used to find and balance the multi-objective conflicts; the Pareto optimal solution of multi-objective optimization is obtained to obtain the optimal combination of process parameters.

9. The intelligent optimization method for the radiation cross-linking process of flexible armor wire according to claim 1 is characterized in that: The S5 is specifically: Apply the optimal process parameters obtained through the optimization algorithm to the production line; Install a real-time data acquisition system on the production line to capture performance test data of processed materials, including cross-linking density, flexibility, and fire resistance; Compare the test data with the production target value to determine whether the test indicators meet the process requirements; If the inspection data deviates from the target, the process parameters are adjusted through the genetic algorithm and applied to the production line again to form a closed loop until the production target is achieved.

10. An intelligent optimization system for the radiation cross-linking process of flexible armor wires, characterized in that: The method comprises a processor, a memory and a computer program stored in the memory. When the processor executes the computer program, the method specifically performs the steps in the intelligent optimization method for the radiation cross-linking process of flexible armor wire as described in any one of claims 1 to 9.

Images