Multi-physics field optimization method and system for micro-nano spraying path

Abstract

The invention relates to the technical field of micro-nano spraying, in particular to a multi-physical field optimization method and system for a micro-nano spraying path, and the method comprises the steps: building a spraying particle motion model according to the particle behavior in the spraying process; carrying out coupling analysis on the spraying particle motion model; current environment parameters and spraying process data are collected, an ideal spraying path is obtained, and the ideal spraying path and the current environment parameters are input into the coupling analysis model to obtain an initial spraying path; optimizing the initial spraying path by using a multi-target particle swarm optimization algorithm based on navigation variables to obtain an optimal spraying path; and carrying out adaptive adjustment on the optimal spraying path by using a multi-task Bayesian federation learning algorithm. According to the method, the physical field is comprehensively considered, the NMOPSO algorithm is adopted to plan the micro-nano spraying path, the optimal spraying path is found, the safe and efficient optimal path can be effectively found in the complex spraying environment, and the spraying efficiency and quality are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of micro-nano spraying, and particularly to a multi-physical field optimization method and system for micro-nano spraying paths. Background Art

[0002] Micro-nano spraying technology is an advanced process for forming micro-nano coatings on the surface of objects. Such coatings have various excellent properties, such as high hardness, high corrosion resistance, good optical properties, etc., and are widely used in fields such as aerospace, automotive manufacturing, and electronic devices. The micro-nano spraying process involves the interaction of multiple physical fields, such as the airflow field, temperature field, stress field, etc. The complex coupling relationship of these physical fields has an important impact on the spraying quality and efficiency. Therefore, it is necessary to adopt a multi-physical field optimization method and system to improve the accuracy and efficiency of micro-nano spraying.

[0003] When planning the micro-nano spraying path, it is necessary to comprehensively consider the influence of multiple physical fields such as the airflow field and temperature field. This requires the spraying equipment to have high-precision control and monitoring capabilities, and be able to monitor and adjust various physical parameters during the spraying process in real time. At the same time, the operator also needs to have rich experience and skills, and be able to flexibly adjust the spraying parameters and path planning according to the actual situation to ensure that the spraying effect reaches the best. The influence of multiple physical fields is an important factor to be considered in the micro-nano spraying path planning. By reasonably controlling the distribution and intensity of physical fields such as the airflow field and temperature field, it is ensured that the coating material can be evenly and stably deposited on the surface of the substrate, thereby obtaining a high-quality coating. At the same time, this also requires the spraying equipment and operators to have higher technical levels and capabilities to adapt to the continuous development and application requirements of micro-nano spraying technology.

[0004] Traditional methods often focus on the consideration of a single physical field and ignore the interaction and mutual influence between multiple physical fields. This isolated analysis method is difficult to accurately capture the complex physical phenomena during the spraying process, resulting in limited optimization effects. Traditional methods often ignore the real-time monitoring and continuous improvement during the spraying process. Once the spraying process is determined, it is very difficult to flexibly adjust and optimize according to the actual situation. This rigid management method not only limits the improvement of spraying efficiency and coating quality, but also may increase production costs and environmental risks. With the progress of technology and the improvement of environmental protection requirements, micro-nano spraying is facing more and more complex and changeable spraying requirements. Due to the lack of flexibility and adaptability, traditional methods often have difficulty in coping with these challenges. Summary of the Invention

[0005] The present invention provides a multi-physical field optimization method and system for micro-nano spraying paths to solve the defect that the existing technology focuses on the consideration of a single physical field and ignores the interaction and mutual influence between multiple physical fields, resulting in limited optimization effects.

[0006] The present invention provides a multi-physical field optimization method for a micro-nano spraying path, including:

[0007] Establish a spraying particle motion model based on the particle behavior during the spraying process.

[0008] Conduct a coupling analysis on the spraying particle motion model to obtain a coupling analysis model.

[0009] Collect the current environmental parameters and spraying process data, obtain an ideal spraying path, where the ideal spraying path includes multiple ideal spraying points and the ideal spraying parameter combinations corresponding to each ideal spraying point, and input the ideal spraying path and the current environmental parameters into the coupling analysis model to obtain an initial spraying path.

[0010] Use a multi-objective particle swarm optimization algorithm based on navigation variables to optimize the initial spraying path to obtain an optimal spraying path.

[0011] Use a multi-task Bayesian federated learning algorithm to adaptively adjust the optimal spraying path.

[0012] According to the multi-physical field optimization method for a micro-nano spraying path provided by the present invention, the spraying particle motion model includes:

[0013] A temperature field unit for generating a temperature distribution map based on the heat generated by the spray gun and the heat exchange between the spraying particles and the surrounding environment.

[0014] An air flow field unit for generating an air flow distribution map based on the air flow changes in the environment where the spray gun is located.

[0015] A gravity field unit for calculating the falling speed of the spraying particles in the gravity field.

[0016] According to the multi-physical field optimization method for a micro-nano spraying path provided by the present invention, the gravity field unit includes:

[0017] An electrostatic field sub-unit for analyzing the interference degree of the electrostatic field on the gravity field and correcting the falling speed of the spraying particles according to the interference degree.

[0018] According to the multi-physical field optimization method for a micro-nano spraying path provided by the present invention, the coupling analysis includes:

[0019] Determine the spraying constraint conditions, where the spraying constraint conditions include: the size and shape of the spraying area, the initial velocity, temperature, and charge state of the spraying particles.

[0020] Establish a change equation for the spraying particle motion model according to the spraying constraint conditions.

[0021] Calculate the solution of the change equation of the spraying particle motion model using particle velocity, particle diameter and mass, gas flow rate and pressure, temperature, time step, and spraying constraint conditions to obtain the coupling relationship of the elements in the spraying particle motion model.

[0022] According to a multi-physical field optimization method for micro-nano spraying paths provided by the present invention, the multi-objective particle swarm optimization algorithm based on navigation variables includes:

[0023] Obtain the initial spraying path data, and determine the initial position of the initial spraying path in the spraying scheme group, the key node distribution and change trend of the initial spraying path.

[0024] Calculate and compare the objective function values of the spraying schemes, and use the navigation variables to adjust the key node distribution and change trend of the spraying path according to the update formula.

[0025] Judge whether the objective function value of the spraying scheme converges. If so, output the optimal spraying path; otherwise, continue to adjust the key node distribution and change trend of the spraying path.

[0026] According to a multi-physical field optimization method for micro-nano spraying paths provided by the present invention, using the navigation variables to adjust the key node distribution and change trend of the spraying path according to the update formula includes:

[0027] Calculate the objective function value of each spraying scheme.

[0028] Compare the objective function values of each spraying scheme, and determine the path situation with the best comprehensive effect of the spraying path in the spraying scheme group and the path situation with the best comprehensive effect among many spraying path schemes.

[0029] Use the navigation variables to update the key node distribution and change trend of the spraying path, change the change trend of the spraying path according to the preset change trend update formula, and adjust the key node distribution of the spraying path according to the preset key node distribution update formula using the updated node change trend.

[0030] According to a multi-physical field optimization method for micro-nano spraying paths provided by the present invention, the calculation formula for the objective function value of each spraying scheme is:

[0031] F = ω1f1 + ω2f2 + ω3f3

[0032] In the formula, f1 is the function of spraying uniformity, f2 is the path length function, f3 is the energy consumption function, and ω1, ω2, and ω3 are the weights corresponding to the objective functions of spraying uniformity, path length, and energy consumption, respectively.

[0033] According to a multi-physical field optimization method for micro-nano spraying paths provided by the present invention, the adaptive adjustment of the optimal spraying path includes:

[0034] Initialize the multi-output Gaussian process model 1 using a local prior distribution.

[0035] Locally train the multi-output Gaussian process model 1 and iteratively adjust the spraying path.

[0036] Repeat the training iteration process to obtain an adaptively adjusted spraying path.

[0037] According to a multi-physical field optimization method for a micro-nano spraying path provided by the present invention, locally training the multi-output Gaussian process model 1 includes:

[0038] Train the multi-output Gaussian process model 1 using spraying process data to capture the relationship between the path width and path overlap rate and the input variables, and obtain the posterior distribution 1 of the path width and overlap rate.

[0039] Upload the posterior distribution 1 on the local device to the global processor and merge it to update the prior distribution 1 of the global multi-output Gaussian process model 1, obtaining the global prior distribution 2.

[0040] Send the global prior distribution 2 back to the local device as the starting point for the next round of training.

[0041] Initialize the multi-output Gaussian process model 1 again using the global prior distribution 2 to obtain the multi-output Gaussian process model 2.

[0042] Train the multi-output Gaussian process model 2 again using spraying process data to obtain the posterior distribution 2 of the path width and overlap rate.

[0043] Use the posterior distribution 2 of the path width and overlap rate as the input for adaptively adjusting the spraying path, and dynamically adjust the width and overlap rate of the spraying path in combination with real-time spraying conditions.

[0044] According to a multi-physical field optimization system for a micro-nano spraying path provided by the present invention, it includes:

[0045] A model establishment module for establishing a spraying particle motion model according to the particle behavior during the spraying process.

[0046] A coupling analysis module for performing coupling analysis on the spraying particle motion model to obtain a coupling analysis model.

[0047] A strategy optimization module for formulating an optimization strategy according to the coupling analysis model to plan the spraying and spraying path.

[0048] An initial path module for collecting the current environmental parameters and spraying process data, obtaining an ideal spraying path, and inputting the ideal spraying path and the current environmental parameters into the coupling analysis model to obtain an initial spraying path.

[0049] The spraying path module is used to optimize the initial spraying path using a multi-objective particle swarm optimization algorithm based on navigation variables to obtain the optimal spraying path.

[0050] The path adaptive adjustment module is used to adaptively adjust the optimal spraying path using a multi-task Bayesian federated learning algorithm.

[0051] A multi-physical field optimization method and system for micro-nano spraying paths provided by the present invention. By comprehensively considering multiple physical fields and using the NMOPSO algorithm to plan the micro-nano spraying path and find the optimal spraying path, it solves the defect in the prior art that the consideration of a single physical field ignores the interaction and mutual influence between multiple physical fields, resulting in limited optimization effects. The beneficial effects obtained are as follows:

[0052] By comprehensively considering multiple physical field models and their interactions, the present invention more accurately simulates the motion behavior of spraying particles, including key indicators such as particle velocity, trajectory, deposition efficiency, and uniformity. The accurate physical field model helps to predict various phenomena during the spraying process, thereby guiding experimental design and parameter optimization. Based on the optimization strategy formulated by the coupling analysis model, the spraying parameters are systematically adjusted to maximize the spraying effect. Through automated and intelligent optimization algorithms, the number of experiments is significantly reduced, and the efficiency and accuracy of parameter optimization are improved. The introduction of path planning technology makes the spraying path more accurate and efficient, further improving the spraying efficiency and quality.

[0053] Using a multi-objective particle swarm optimization algorithm based on navigation variables to plan the micro-nano spraying path and find the optimal spraying path. Through multi-objective optimization technology, the NMOPSO algorithm can effectively find a safe and efficient optimal path in a complex spraying environment, improving the spraying efficiency and quality. The optimal path planning ensures the uniform distribution of spraying particles on the object surface, reduces spraying overlap and omission areas, improves spraying uniformity and coverage rate. By optimizing the spraying path, unnecessary spraying actions and repeated spraying are reduced, thereby reducing material consumption and costs. At the same time, the optimal path planning also reduces energy consumption and emissions during the spraying process, meeting the requirements of green manufacturing and sustainable development. By adjusting the algorithm parameters and constraint conditions, it is convenient to adapt to different spraying requirements and scenarios, improving the versatility and practicality of the spraying system. The optimal path planning is an important part of the intelligence and automation of spraying technology, which helps to improve production efficiency and product quality, reduce manual intervention and costs, and promote the development of spraying technology to a higher level.

[0054] The finite volume field-circuit coupling algorithm is used to simulate the spraying parameter combinations to obtain the optimal parameter combinations, more accurately simulate the interaction and influence of multiple physical fields during the spraying process, more precisely calculate parameters such as the movement trajectory, velocity, and temperature of the spraying particles, as well as the distribution of physical fields such as the electric field and magnetic field during the spraying process. This helps to improve the accuracy of the simulation results and provide a reliable basis for optimizing the spraying parameters. By simulating the spraying effects under different spraying parameter combinations, the influence of various parameters on the spraying quality is systematically evaluated. The finite volume field-circuit coupling algorithm can comprehensively consider the effects of multiple physical fields and find the optimal parameter combination that meets specific spraying requirements. This helps to improve the spraying efficiency, reduce costs, improve the spraying quality, and meet the requirements of specific application scenarios. Traditional methods for optimizing spraying parameters often rely on a large number of experiments and trial-and-error, which are time-consuming, laborious, and costly. By using the finite volume field-circuit coupling algorithm for simulation, the effects of different parameter combinations can be quickly evaluated on the computer, reducing the number of experiments and costs. This helps to shorten the R & D cycle and accelerate the promotion and application of new technologies. The finite volume field-circuit coupling algorithm has strong adaptability and can handle the spraying problems of objects with different shapes, sizes, and materials.

[0055] By adjusting the algorithm parameters and boundary conditions, it can easily adapt to different spraying requirements and scenarios, improving the versatility and practicality of the spraying system. Using the finite volume field-circuit coupling algorithm for optimizing spraying parameters helps to promote the innovation and development of spraying technology. By continuously optimizing the spraying parameter combinations, spraying processes and products with higher performance, lower costs, and more environmental protection can be developed. This helps to improve the technical level and market competitiveness of enterprises and promote the development of spraying technology to a higher level.

[0056] The spraying technology based on the physical field model and optimization strategy represents the forefront research direction in the field of micro-nano spraying. Through technological innovation, spraying processes and products with higher performance, lower costs, and more environmental protection can be developed, enhancing the market competitiveness and brand influence of enterprises. Brief Description of the Drawings

[0057] To more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings without creative efforts.

[0058] Figure 1 It is a flowchart of a multi-physical field optimization method for a micro-nano spraying path provided by an embodiment of the present invention;

[0059] Figure 2It is a flowchart for adaptively adjusting the optimal spraying path using the multi-task Bayesian federated learning algorithm provided by an embodiment of the present invention;

[0060] Figure 3 It is a module diagram of a multi-physical field optimization method for a micro-nano spraying path provided by an embodiment of the present invention. Specific embodiments

[0061] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without making creative efforts fall within the protection scope of the present invention.

[0062] The following will be combined with Figures 1 - 3 Describe a multi-physical field optimization method and system for a micro-nano spraying path of the present invention.

[0063] As Figure 1 shown, a multi-physical field optimization method for a micro-nano spraying path provided by an embodiment of the present invention includes:

[0064] Establish a spraying particle motion model based on the particle behavior during the spraying process. The spraying particle motion model includes multiple physical field models. The spraying particle motion model is the basis for describing the particle behavior during the spraying process. By establishing the motion equation of the particle, the velocity, acceleration, and trajectory of the particle can be simulated. During the spraying process, the particles are ejected from the spray gun and are affected by various forces such as temperature, air flow, gravity, and electric field force. The spraying particle motion model comprehensively considers the effects of these forces and the interactions between the particles and the air and the substrate surface to establish each physical field model.

[0065] The temperature field model is based on the heat transfer and distribution during the spraying process. During the spraying process, heat will spread in the spraying area through conduction, convection, and radiation, affecting the quality and performance of the coating. Establishing the temperature field model can predict the temperature change and the heat affected area in the spraying area, providing an important basis for optimizing the spraying process. Describing the spatial distribution of the temperature in the spraying area is mainly determined by the heat generated by the spray gun and the heat exchange between the particles and the surrounding environment. The change of the temperature field will affect the velocity and direction of the particles. The temperature gradient will generate thermophoretic force, causing the particles to move towards the area with lower temperature. In plasma spraying, the high-temperature plasma arc is the main heat source, enabling the spraying material to be quickly heated to the molten or semi-molten state. Heat is transferred to the spraying particles through heat conduction and heat radiation, and at the same time, the particles will also dissipate heat to the surrounding environment, resulting in a complex distribution of the temperature field in the spraying area.

[0066] The airflow field model analyzes the airflow field during the spraying process, which is mainly determined by the airflow generated by the spray gun and the airflow in the surrounding environment. The magnitude and direction of the airflow velocity determine the acceleration degree and flight trajectory of the particles. In plasma spraying, the airflow velocity is usually very high, reaching hundreds of meters per second. The stability of the airflow has an important impact on the spraying quality. Unstable airflow will lead to the deviation of particle trajectories and uneven coatings. At the same time, there is an interaction between the airflow and the particles. The airflow exerts a resistance on the particles, and at the same time, the particles will also change the flow state of the airflow. The spraying gas plays a key role in the spraying process, and its flow velocity and pressure distribution will affect the trajectories and distributions of the spraying particles. By establishing a flow field model, the influence of the airflow on the spraying effect can be evaluated, providing a basis for optimizing the spraying path and parameters.

[0067] The gravitational field is the attraction field of the earth on objects, which also has a certain impact on the movement of particles during the spraying process. The gravitational acceleration is a constant, which determines the falling speed of the particles in the gravitational field. For smaller particles, the influence of gravity is relatively small. For larger particles, gravity will significantly affect their movement trajectories. The spraying angle will also affect the degree of influence of gravity. The smaller the spraying angle, the greater the influence of gravity on the movement of particles. It also includes an electrostatic field, which is used to analyze the interference degree of the electrostatic field on the gravitational field and correct the falling speed of the spraying particles according to the interference degree.

[0068] Perform a coupling analysis on the interactions between the various physical field models to obtain a coupling analysis model. The coupling analysis is for the interactions between the various physical fields, and the interaction mechanism will affect the spraying effect and coating quality.

[0069] The specific steps of the coupling analysis are as follows:

[0070] It is necessary to define the problem domain and boundary conditions to clarify the physical space and conditions for analysis. In the spraying particle motion model, this includes determining the size and shape of the spraying area, as well as the initial velocity, temperature, and charge state of the spraying particles, etc. At the same time, boundary conditions such as the airflow velocity, temperature distribution, and electric field strength in the spraying environment also need to be considered.

[0071] Secondly, select appropriate physical models to describe the changes in each physical field. For the flow field, use the fluid dynamics equations to describe the changes in parameters such as the velocity, direction, and density of the airflow. For the temperature field, use the heat conduction equation or heat convection equation to describe the temperature distribution and changes. For the electric field force field, use the electric potential equation or electric field strength equation to describe the changes in the electric field. The selection of these physical models should be based on the specific characteristics and accuracy requirements of the problem.

[0072] Use the finite element algorithm to design the coupling mechanism between different physical fields. In the spray particle motion model, the interaction between various physical fields is complex and requires an appropriate coupling mechanism to describe. For example, the change in the flow field may affect the distribution of the temperature field because the flow of air will carry away or bring heat. Similarly, the electric field and force field may also affect the movement of particles in the flow field by changing the charged state of the particles. Therefore, additional algorithms and interfaces are needed to achieve the coupling between these physical fields.

[0073] Select numerical methods and mesh generation strategies. The accuracy and efficiency of mesh generation have an important impact on the solution results. Too coarse a mesh may lead to inaccurate solution results, while too fine a mesh will increase the computational amount and reduce the solution efficiency.

[0074] The change in the temperature field will affect the speed and direction of the spray particles. The high-temperature area accelerates the particle movement, and the low-temperature area decelerates or even changes the direction of the particles. At the same time, the thermophoretic force generated by the temperature gradient will also guide the particles to move towards the low-temperature area, resulting in uneven coating distribution.

[0075] The speed and direction of the air flow in the air flow field will change the flight path of the particles, increase scattering and diffusion, and affect the uniformity of the coating. The interaction between the particles and the air flow will generate resistance, further affecting the motion state of the particles.

[0076] In electrostatic spraying, the electric field makes the spray particles charged and subject to the action of the electric field force. The charge amount and charge sign of the particles determine the direction and magnitude of the force. The uneven distribution of the electric field intensity leads to changes in the movement trajectory and distribution of the particles. There may also be an interaction between the electric field and the air flow field, jointly affecting the movement of the particles.

[0077] These coupling relationships need to be precisely controlled and optimized during the spraying process. Use numerical simulation technology to analyze the multi-physical field coupling. By using numerical simulation methods such as computational fluid dynamics and computational electromagnetics, conduct coupling simulation of multi-physical fields to predict the physical phenomena and results during the spraying process. Numerical simulation provides detailed physical field distribution and change information, providing an important basis for optimizing the spraying process and parameters.

[0078] Use the multi-objective particle swarm optimization algorithm based on navigation variables to plan the micro-nano spraying path and find the optimal spraying path. Path planning optimizes the spraying path according to the movement trajectory of spraying particles and the temperature and flow field distribution. By reasonably planning the spraying path, the coating uniformity, quality stability can be ensured, and the energy consumption and emissions during the spraying process can be reduced. When formulating optimization strategies, traditional methods often rely on empirical judgment and trial-and-error methods, lacking scientific data support and theoretical guidance. This method is not only inefficient but also difficult to ensure the accuracy and reliability of the optimization results. The NMOPSO algorithm is an advanced algorithm specifically used to handle high-dimensional multi-objective optimization problems. The algorithm models the path planning problem as a multi-objective optimization problem and uses particle swarm optimization technology to find the Pareto optimal path that meets specific constraints. The core lies in introducing navigation variables to represent the path, so as to better consider the constraints of the problem, and generating a set of non-dominated solutions through multi-objective optimization methods to meet different application requirements.

[0079] The NMOPSO algorithm can handle multiple objective functions simultaneously, which is particularly important in complex application scenarios such as micro-nano spraying. Because the spraying effect is often affected by multiple factors, such as coating uniformity, spraying efficiency, energy consumption, etc. The algorithm can find a balance among these objectives and generate a set of solutions that meet multiple objectives. The algorithm can consider complex constraints, such as physical field constraints and robot kinematic constraints during the spraying process. By introducing navigation variables, the algorithm can better represent and constrain the solution space of the optimization problem, ensuring that the generated path not only meets the actual requirements but also conforms to physical laws. The NMOPSO algorithm has global search ability and can quickly converge to the optimal solution, and can find the optimal path that meets the constraint conditions in a relatively short time. This is of great significance for improving the efficiency and accuracy of micro-nano spraying.

[0080] The specific steps of the NMOPSO algorithm are as follows:

[0081] Sort out the data related to the initial spraying path, and determine the initial position of the initial spraying path in the spraying scheme group, the distribution and change trend of the key nodes of the initial spraying path.

[0082] Calculate the objective function value of each spraying scheme.

[0083] Compare the objective function values of each spraying scheme, and determine the path situation with the best comprehensive effect in the spraying scheme group and the path situation with the best comprehensive effect among many spraying path schemes.

[0084] Update the key node distribution and change trend of the spraying path using navigation variables, change the change trend of the spraying path according to the preset change trend update formula, and adjust the key node distribution of the spraying path according to the preset key node distribution update formula using the updated node change trend, so that the spraying path moves in the direction more likely to find the optimal path.

[0085] Repeat the steps of calculating the objective function value, determining the extreme value, and updating the particle velocity and position above, and continuously iterate. Each iteration further explores a better spraying path based on the previous one. Judge whether the objective function value of the spraying scheme converges. Otherwise, continue the iteration. If it converges, output the optimal spraying path.

[0086] The objective function value of each spraying scheme includes a spraying uniformity objective function, a spraying path length objective function, and a spraying energy consumption objective function.

[0087] The formula expression of the spraying uniformity objective function is:

[0088]

[0089] In the formula, n is the sampling point, n takes 1, 2...i, ti is the actual spraying thickness obtained at the i-th sampling point, is the ideal average spraying thickness.

[0090] The calculation formula of the spraying path length is:

[0091]

[0092] In the formula, (x j , y j ) are the node coordinates, j = 1, 2,..., m, and m is the key node in the spraying path.

[0093] This value reflects the length of the equipment movement distance during the spraying process. Naturally, the smaller it is, the better, which means a shorter working stroke and can save time and resources, etc.

[0094] The formula expression of the spraying energy consumption objective function is:

[0095]

[0096] In the formula, v is the equipment running speed, P is the power of the spraying equipment, and f2 is the path length.

[0097] The comprehensive objective function calculation formula is:

[0098] F = ω1f1 + ω2f2 + ω3f3

[0099] Where f1 is the function of spraying uniformity, f2 is the path length function, f3 is the energy consumption function, and ω1, ω2, and ω3 are the weights corresponding to the spraying uniformity, path length, and energy consumption objective functions, respectively.

[0100] As Figure 2 shown, the specific steps for adaptively adjusting the optimal spraying path using the multi-task Bayesian federated learning algorithm are as follows:

[0101] On each local device, initialize the multi-output Gaussian process model 1 using the local prior distribution 1. This prior distribution may be based on historical data, expert knowledge, or default settings.

[0102] Use the spraying process data to train the multi-output Gaussian process model 1 to capture the relationship between the path width and path overlap rate and the input variables, and obtain the posterior distribution 1 of the path width and overlap rate.

[0103] Upload the posterior distributions on each local device to the global processor. These posterior distributions contain the latest information about the path width and overlap rate. Information aggregation and update of the global prior distribution on the global processor: On the global processor, use aggregation algorithms such as Bayesian averaging and Bayesian fusion to merge these posterior distributions 1 to update the global MOGP prior distribution 1 and obtain the global prior distribution 2. The updated global prior distribution reflects the integration of information and knowledge from all local devices.

[0104] Send the global prior distribution 2 back to the local device as the starting point for the next round of training. This step ensures that each local device has the latest global information, enabling more accurate predictions in the next round of training.

[0105] Use the global prior distribution 2 to initialize the multi-output Gaussian process model 1 again to obtain the multi-output Gaussian process model 2. Use the spraying process data to train the multi-output Gaussian process model 2 again to obtain the posterior distribution 2 of the path width and overlap rate.

[0106] Use the posterior distribution 2 as the input for the next step of adaptively adjusting the spraying path. Use the information in the new posterior distribution, such as the mean and variance, as inputs, and combine real-time spraying conditions, such as the workpiece shape and material type, to dynamically adjust the width and overlap rate of the spraying path. The goal of adaptive adjustment is to ensure the consistency of spraying quality while improving spraying efficiency.

[0107] Through continuous iterative optimization, the width and overlap rate of the spraying path gradually approach the optimal values. In each iteration, use the output of the previous step as the input for the next step to ensure the continuity and accuracy of information.

[0108] A multi - physical - field optimization method and system for micro - nano spraying paths provided by the present invention. By comprehensively considering multiple physical fields and using the NMOPSO algorithm to plan the micro - nano spraying paths, the optimal spraying path is found, solving the defect in the prior art that the consideration focuses on a single physical field and ignores the interaction and mutual influence between multiple physical fields, resulting in limited optimization effect. The beneficial effects obtained are as follows:

[0109] By comprehensively considering multiple physical - field models and their interactions, the present invention more accurately simulates the motion behavior of spraying particles, including key indicators such as particle velocity, trajectory, deposition efficiency, and uniformity. The accurate physical - field model helps to predict various phenomena during the spraying process, thereby guiding experimental design and parameter optimization. Based on the optimization strategy formulated by the coupled - analysis model, the spraying parameters are systematically adjusted to maximize the spraying effect. Through automated and intelligent optimization algorithms, the number of experiments is significantly reduced, and the efficiency and accuracy of parameter optimization are improved. The introduction of path - planning technology makes the spraying path more accurate and efficient, further improving the spraying efficiency and quality.

[0110] As Figure 3 shown, a multi - physical - field optimization system for micro - nano spraying paths includes:

[0111] A model - building module for establishing a spraying - particle motion model according to the particle behavior during the spraying process.

[0112] A coupling - analysis module for performing coupling analysis on the spraying - particle motion model to obtain a coupled - analysis model.

[0113] A strategy - optimization module for formulating an optimization strategy according to the coupled - analysis model to plan the spraying and spraying paths.

[0114] A spraying - parameter unit for simulating the spraying - parameter combination using the finite - volume field - circuit coupling algorithm to obtain the relatively optimal parameter combination.

[0115] A spraying - path unit for planning the micro - nano spraying path using the multi - objective particle - swarm optimization algorithm based on navigation variables to find the relatively optimal spraying path.

[0116] An experimental - verification module for experimentally verifying the relatively optimal parameter combination and the optimal path to obtain experimental results.

[0117] A strategy - correction module for correcting and improving the physical - field model and the optimization strategy according to the experimental results to obtain an improved spraying - particle motion model.

[0118] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, i.e., located in one place or distributed to multiple network units. Select some or all of the modules according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative effort.

[0119] Through the description of the above embodiments, those skilled in the art clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course also by hardware. Based on such an understanding, the above technical solution, in essence, or the part that contributes to the prior art is embodied in the form of a software product. The computer software product is stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (personal computer, server, or network device, etc.) to execute the methods of each embodiment or some parts of the embodiments.

[0120] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments or perform equivalent replacements for some of the technical features. These modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multi-physical field optimization method for micro-nano spraying paths, characterized in that, Including: Establish a spraying particle motion model based on the particle behavior during the spraying process; Conduct a coupling analysis on the spraying particle motion model to obtain a coupling analysis model; Collect the current environmental parameters and spraying process data, obtain an ideal spraying path, where the ideal spraying path includes multiple ideal spraying points and the corresponding ideal spraying parameter combinations for each ideal spraying point, and input the ideal spraying path and the current environmental parameters into the coupling analysis model to obtain an initial spraying path; Optimize the initial spraying path using a multi-objective particle swarm optimization algorithm based on navigation variables to obtain an optimal spraying path; Use a multi-task Bayesian federated learning algorithm to adaptively adjust the optimal spraying path.

2. The multi-physical field optimization method for the micro-nano spraying path according to claim 1, wherein The spraying particle motion model includes: A temperature field unit for generating a temperature distribution map based on the heat generated by the spray gun and the heat exchange between the spraying particles and the surrounding environment; An airflow field unit for generating an airflow distribution map based on the airflow changes in the environment where the spray gun is located; A gravity field unit for calculating the falling speed of the spraying particles in the gravity field.

3. A multi-physical field optimization method for a micro-nano spraying path according to claim 1, characterized in that The gravity field unit includes: An electrostatic field sub-unit for analyzing the interference degree of the electrostatic field on the gravity field and correcting the falling speed of the spraying particles according to the interference degree.

4. A multi-physical field optimization method for a micro-nano spraying path according to claim 1, characterized in that, The coupling analysis includes: Determine the spraying constraint conditions, where the spraying constraint conditions include: the size and shape of the spraying area, the initial velocity, temperature, and charge state of the spraying particles; Establish a change equation for the spraying particle motion model according to the spraying constraint conditions; Calculate the solution of the change equation of the spraying particle motion model using particle velocity, particle diameter and mass, gas flow rate and pressure, temperature, time step, and spraying constraint conditions to obtain the coupling relationship of the units in the spraying particle motion model.

5. A multi-physical field optimization method for a micro-nano spraying path according to claim 1, characterized in that The multi-objective particle swarm optimization algorithm based on navigation variables includes: Obtain the initial spraying path data, and determine the initial position of the initial spraying path in the spraying scheme group, the key node distribution and change trend of the initial spraying path; Calculate and compare the objective function values of the spraying schemes, and use the navigation variables and adjust the key node distribution and change trend of the spraying path according to the update formula; Judge whether the objective function value of the spraying scheme converges. If so, output the optimal spraying path, otherwise continue to adjust the key node distribution and change trend of the spraying path.

6. The multi-physical field optimization method for a micro-nano spraying path according to claim 5, wherein Adjusting the key node distribution and change trend of the spraying path using the navigation variable according to the update formula includes: Calculate the objective function value of each spraying scheme; Compare the objective function values of each spraying scheme, and determine the path situation with the best comprehensive effect of the spraying path in the spraying scheme group and the path situation with the best comprehensive effect among many spraying path schemes; Use the navigation variable to update the key node distribution and change trend of the spraying path, change the change trend of the spraying path according to the preset change trend update formula, and adjust the key node distribution of the spraying path according to the preset key node distribution update formula using the updated node change trend.

7. A multi-physical field optimization method for a micro-nano spraying path according to claim 5, characterized in that The calculation formula for the objective function value of each spraying scheme is: F = ω1f1 + ω2f2 + ω3f3 In the formula, f1 is the function of spraying uniformity, f2 is the path length function, f3 is the energy consumption function, and ω1, ω2, and ω3 are the weights corresponding to the spraying uniformity, path length, and energy consumption objective functions, respectively.

8. The multi-physical field optimization method for the micro-nano spraying path according to claim 1, characterized in that The adaptive adjustment of the optimal spraying path includes: Using the local prior distribution to initialize the multi-output Gaussian process model 1; Locally training the multi-output Gaussian process model 1 and iteratively adjusting the spraying path; Repeating the training and iteration process to obtain the adaptively adjusted spraying path.

9. The multi-physical field optimization method for the micro-nano spraying path according to claim 1, characterized in that, The local training of the multi-output Gaussian process model 1 includes: Using the spraying process data to train the multi-output Gaussian process model 1, capturing the relationship between the path width and path overlap rate and the input variables, and obtaining the posterior distribution 1 of the path width and overlap rate; Uploading the posterior distribution 1 on the local device to the global processor for merging, updating the prior distribution 1 of the global multi-output Gaussian process model 1, and obtaining the global prior distribution 2; Sending the global prior distribution 2 back to the local device as the starting point for the next round of training; Using the global prior distribution 2 to re-initialize the multi-output Gaussian process model 1 to obtain the multi-output Gaussian process model 2; Using the spraying process data to retrain the multi-output Gaussian process model 2 to obtain the posterior distribution 2 of the path width and overlap rate; Using the posterior distribution 2 of the path width and overlap rate as the input for the adaptive adjustment of the spraying path, and dynamically adjusting the width and overlap rate of the spraying path in combination with the real-time spraying conditions.

10. A multi-physical field optimization system for a micro-nano spraying path, which adopts a multi-physical field optimization method for a micro-nano spraying path as described in any one of claims 1 to 9, characterized in that, It includes: A model establishment module for establishing a spraying particle motion model according to the particle behavior during the spraying process; A coupling analysis module for performing coupling analysis on the spraying particle motion model to obtain a coupling analysis model; A strategy optimization module for formulating an optimization strategy according to the coupling analysis model to plan the spraying and spraying path; An initial path module for collecting the current environmental parameters and spraying process data, obtaining the ideal spraying path, and inputting the ideal spraying path and the current environmental parameters into the coupling analysis model to obtain the initial spraying path; A spraying path module for optimizing the initial spraying path using a multi-objective particle swarm optimization algorithm based on navigation variables to obtain the optimal spraying path; A path adaptive adjustment module for adaptively adjusting the optimal spraying path using a multi-task Bayesian federated learning algorithm.

Images