A method, a control system, and a method for improving the forming accuracy of polymer laser powder bed melting forming.

Abstract

This invention discloses a method, system, and approach for controlling the precision of polymer laser powder bed fusion molding (LPBF), belonging to the field of additive manufacturing technology. It includes: acquiring temperature distribution data during the polymer LPBF process; constructing a database based on the temperature distribution data and the process parameters used for printing the current layer; establishing a precise control model for the polymer LPBF process based on machine learning; training the precise control model using the database, establishing a mapping relationship between process parameters, printed layer thickness, and layer-by-layer temperature field distribution, obtaining a negative feedback mechanism between process parameters and the real-time temperature field of the printed layer, and achieving precise control of the polymer laser powder bed fusion molding precision. This invention utilizes the superiority of reinforcement learning algorithms in big data processing to improve the heat accumulation phenomenon during the molding process, eliminate defects such as warping and agglomeration, and has guiding significance for improving the precision and performance of LPBF-processed parts.

Description

Technical Field

[0001] This invention belongs to the field of additive manufacturing technology and relates to a method, control system and method for improving the accuracy of polymer laser powder bed melting forming. Specifically, it relates to a method based on machine learning to improve the heat accumulation effect in the polymer material laser powder bed melting process, reduce thermal warpage deformation and improve forming accuracy. Background Technology

[0002] Laser Powder Bed Fusion (LPBF) is an advanced additive manufacturing method based on the principle of powder bed melting. A powder feeding mechanism uniformly lays a layer of powder material, such as metal or polymer, on a forming platform. A high-power laser scans and melts the powder according to a software-planned path. The molten powder solidifies during cooling. The powder bed then moves downwards by one layer thickness, the powder feeding mechanism re-lays powder, and the laser scanning, melting, and solidification processes are repeated until a three-dimensional component is formed. Compared to other manufacturing technologies, LPBF offers several advantages, including high forming efficiency, excellent part strength, and a wide range of applicable materials. It can also manufacture parts with complex shapes without requiring additional support structures. Therefore, it is widely used in aerospace, automotive, and medical fields, providing new solutions for rapid, flexible, and efficient manufacturing, while also driving the development of customized and innovative designs.

[0003] Polymer materials typically possess linear or cross-linked chemical structures, exhibiting high specific strength, excellent processability, low thermal conductivity, good chemical stability, good electrical insulation, and good decorative properties. With increasing environmental awareness and growing demands for resources and the environment, polymer components fabricated using the LPBF (Liquid Polymer Material Burner) process are gradually becoming one of the most ideal components in various fields, showing broad application prospects in electronics, aerospace, automotive, and biomedicine. However, improving the molding accuracy of polymer LPBF remains a significant challenge. The main reasons for insufficient precision in molded components are defects such as part warping and powder agglomeration that occur during the LPBF process. Among these factors, the real-time temperature field of the printed layer is the most direct influence on warping and agglomeration. The key factors affecting the real-time temperature field are the heat accumulation effect and real-time process parameters: In the initial stage of printing, the laser heats and melts the powder on the surface of the part, increasing the heat of the part and the surrounding powder. As printing progresses, the heat absorbed by the part in the early stages is gradually released. When a new layer of powder is applied, the powder accumulated in the forming cylinder transfers heat to the newly applied powder, causing the temperature of the newly applied powder layer to rise rapidly. Thus, the processed solid part effectively becomes a new heat source for the processed powder—the bottom heat source—leading to heat accumulation. Existing polymer LPBF forming processes cannot achieve adaptive control of process parameters during printing, and the heat accumulation phenomenon cannot be resolved. This results in uneven distribution of the real-time temperature field in the printed layer during the polymer LPBF process, leading to defects such as part warping and powder agglomeration, ultimately resulting in insufficient consistency in the precision and performance of the formed parts.

[0004] Therefore, improving the heat accumulation effect of polymer materials in the LPBF process and achieving adaptive control of real-time process parameters are essential. This can not only effectively improve the accuracy and performance of the molded parts, but also has significant implications for the stable printing and practical application of engineering components made from high-melting-point polymer materials such as PEEK, PEAK, PU, ​​and their composites. However, current methods for controlling heat accumulation are mostly limited to controlling the temperature range of the molding cavity and the uniformity of the preheating temperature field, with insufficient research focusing on the precise control of the real-time temperature field of the printed layer during the molding process. Summary of the Invention

[0005] In order to overcome the shortcomings of the prior art, the present invention aims to provide a method, a control system and a method for controlling the precision of polymer laser powder bed fusion molding, so as to solve the technical problem of high-precision and high-performance printing of polymer LPBF process, which is currently unable to accurately control the real-time temperature field of the printing layer, resulting in defects such as warping and agglomeration during the molding process.

[0006] To achieve the above objectives, the present invention employs the following technical solution:

[0007] This invention discloses a method for controlling the precision of polymer laser powder bed melting, comprising the following steps:

[0008] S1: Obtain temperature distribution data during the polymer LPBF process;

[0009] S2: Build a database based on temperature distribution data and the process parameters used to print the current layer;

[0010] S3: Based on machine learning, establish a precise control model for polymer LPBF process;

[0011] S4: Utilize the database to train a precise control model for the polymer LPBF process, establish the mapping relationship between process parameters, printing layer thickness and layer-by-layer temperature field distribution, obtain the negative feedback mechanism between process parameters and real-time temperature field of the printing layer, and achieve precise control of polymer laser powder bed melting forming accuracy.

[0012] Preferably, in S1, the temperature distribution data includes the printing layer thickness and the temperature distribution status of each pixel in the real-time printing layer of the formed part;

[0013] Based on the difference between the temperature distribution of each pixel on the single-track laser scanning path and the target temperature distribution during the printing process, the laser energy density and lamp power are adjusted in real time. The laser energy density is jointly controlled by the laser power, scanning speed and scanning spacing.

[0014] Preferably, in S2, the process parameters include preheating temperature, lamp power, laser power, scanning speed, and scanning spacing.

[0015] Preferably, in S3, the machine learning uses a reinforcement learning algorithm, and its database consists of multiple episodes, where each episode is defined as the entire process of completing one laser single-channel scan.

[0016] More preferably, in S3, the established precise control model for the polymer LPBF process includes an environmental state function S, a model action range A, an environmental reward R, and a value function Q; the environmental state function S, the model action range A, and the environmental reward R are independent of each other, and the value function Q is related to the first three functions.

[0017] Wherein, the environment state function is ,in, x , y Used to determine the position of each pixel in the real-time printing layer of the formed part during the printing process. z This is the current printed layer thickness. This indicates the real-time temperature distribution of the printed layers on the formed part.

[0018] The model's action range A is defined as a set of process parameters that can be adjusted in real time, including lamp power, laser power, scanning speed, and scanning spacing.

[0019] The environmental reward R is the difference between the temperature distribution of each pixel in the next layer and the target temperature distribution after a single laser scan is completed.

[0020] The value function Q is ,in: This represents the strategy executed by an individual, specifically the temperature distribution state of the printed layer on the molded part. s The probability of adjusting a certain controllable process parameter within the action range A to a is used to explore better parameters. Indicates the execution of an operation a Temperature distribution of the next layer of the post-formed part; For the next state The process parameters that need to be adjusted; To perform the operation a Later in state s The rewards observed in the process; This is a decay factor used to control the weight of the current reward and the rewards for subsequent states; It represents the expected value of the value function, used to measure the magnitude of the theoretical reward and to provide a basis for the model to choose subsequent actions.

[0021] More preferably, by utilizing the reward-maximizing characteristic of the value function, the absolute error between the temperature distribution on a single laser scanning path and the target temperature distribution is summed as a decreasing term of the value function.

[0022] Preferably, in S4, there is a negative feedback mechanism between the process parameters and the real-time temperature field of the printed layer.

[0023] More preferably, the database data is allocated in a 9:1 ratio, with 90% of the data used as the training set and the remaining 10% as the validation set. The Proximal Policy Optimization (PPO) algorithm is used to iteratively optimize the data, leveraging its superior policy selection capabilities to train the model. After training, the model is validated using the validation set to ensure its accuracy. Furthermore, L2 regularization is applied to the model, optimizing the weights of each feature in the environment state function S to prevent overfitting and excessive model complexity.

[0024] Preferably, the polymer comprises one or more of polyaryletherketone, polyetherketone, polyetherketoneketone, polyetheretherketoneketone, polyimide, polycarbonate, polyethylene, polypropylene, polystyrene, and nylon.

[0025] This invention also discloses a control system based on the above-mentioned method for controlling the precision of polymer laser powder bed melting, comprising:

[0026] The data acquisition module is used to acquire temperature distribution data during the polymer LPBF process;

[0027] The database construction module is used to build a database based on temperature distribution data and the process parameters used to print the current layer;

[0028] The control model building module is used to build a precise control model for the polymer LPBF process based on machine learning.

[0029] The model training module is used to train a precise control model for the polymer LPBF process using a database, establish a mapping relationship between process parameters, printing layer thickness and layer-by-layer temperature field distribution, and obtain a negative feedback mechanism between process parameters and real-time temperature field of the printing layer.

[0030] Preferably, the method includes: building a precise control model of the polymer LPBF process on the LPBF equipment, and improving the heat accumulation effect in the polymer LPBF process by adjusting the printing process parameters in real time, reducing the warpage deformation of the polymer material during the forming process, and improving the accuracy of the formed parts.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] This invention discloses a method for controlling the precision of polymer laser powder bed fusion molding (LPBF). On one hand, this invention constructs a database based on the temperature distribution data acquired during the polymer LPBF process and the process parameters adopted for printing the current layer. The database is then used to fully train a precise control model for the polymer LPBF process established during machine learning. The resulting environmental state function more accurately maps the impact of heat accumulation on the molded part itself and the surrounding powder. On the other hand, this invention also considers the improvement effect of process parameters on the forming precision of polymer LPBF. Multiple process parameters are coupled to establish a mapping relationship between process parameters, printing layer thickness, and layer-by-layer temperature field distribution. This obtains a negative feedback mechanism between process parameters and the real-time temperature field of the printing layer, thereby achieving adaptive control of LPBF forming process parameters, improving the heat accumulation effect during the forming process, and ultimately improving the forming precision of polymer materials in the LPBF process.

[0033] Furthermore, the environmental state function in the polymer LPBF process described in this invention is obtained by extracting the temperature distribution of each pixel in the real-time printed layer of the molded part during the process using an infrared thermal imager. Compared with directly extracting the overall temperature inside the molding cavity, the environmental state function obtained by this method can more accurately map the effect of heat accumulation on the molded part itself and the surrounding powder.

[0034] Furthermore, in the database for training models described in this invention, each episode is defined as the entire process of completing one laser single-channel scan. This allows for the efficient collection of a large number of data samples in a single data collection experiment, avoiding the large amount of repetitive and tedious data collection process required by general machine learning.

[0035] Furthermore, this invention considers the proactive exploration and optimization of the strategy, defining a strategy executed by an individual in the model. The model has a probability of employing the optimal strategy. By selecting new strategies within the process window and exploring better possibilities, the model can continuously improve the molding accuracy of polymer LPBF over time.

[0036] Furthermore, this invention considers the overfitting problem that may exist in algorithm learning. A regularization term is introduced into the value function, which is defined as the difference between the lowest and highest temperatures of each pixel observed during an episode. This avoids abnormal strategies that would cause a sudden rise in the temperature of the printed layer of the formed part during the process due to model selection, thereby improving the safety performance of the model.

[0037] Furthermore, the heat received by the real-time printed layer of the formed part during the scanning process is regulated by adaptively adjusting the laser energy density. The laser energy density is jointly controlled by the laser power, scanning speed and scanning spacing. The overall temperature inside the forming cavity is regulated by adaptively adjusting the lamp power. Attached Figure Description

[0038] Figure 1 This is a flowchart of the control method of the present invention;

[0039] Figure 2 This is a flowchart of the method for improving the accuracy of polymer laser powder bed melting and forming according to the present invention;

[0040] Figure 3 This is a schematic diagram illustrating the principle of heat accumulation effect in an embodiment of the present invention;

[0041] Figure 4 This is a scatter plot showing the temperature change inside the forming cavity during the printing process in an embodiment of the present invention.

[0042] Figure 5This is a topographic image of the initial formed part printed in an embodiment of the present invention;

[0043] Figure 6 This is a schematic diagram illustrating the extraction principle of each episode from the database in this embodiment of the invention.

[0044] Figure 7 This is a topographic image of the formed part after printing 10 layers to form the bottom heat source in an embodiment of the present invention;

[0045] Figure 8 This is a topographic image of the final molded part obtained in an embodiment of the present invention;

[0046] Figure 9 This is a topographic image of the final molded part obtained in the comparative example of the present invention. Detailed Implementation

[0047] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0048] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0049] The present invention will now be described in further detail with reference to the accompanying drawings:

[0050] See Figure 1 The present invention discloses a method for controlling the precision of polymer laser powder bed melting, which includes the following steps:

[0051] S1: Obtain temperature distribution data during the polymer LPBF process;

[0052] S2: Build a database based on temperature distribution data and the process parameters used to print the current layer;

[0053] S3: Based on machine learning, establish a precise control model for polymer LPBF process;

[0054] S4: Utilize the database to train a precise control model for the polymer LPBF process, establish the mapping relationship between process parameters, printing layer thickness and layer-by-layer temperature field distribution, obtain the negative feedback mechanism between process parameters and real-time temperature field of the printing layer, and achieve precise control of polymer laser powder bed melting forming accuracy.

[0055] The present invention provides a control system based on the above-mentioned method for controlling the precision of polymer laser powder bed melting, comprising:

[0056] The data acquisition module is used to acquire temperature distribution data during the polymer LPBF process;

[0057] The database construction module is used to build a database based on temperature distribution data and the process parameters used to print the current layer;

[0058] The control model building module is used to build a precise control model for the polymer LPBF process based on machine learning.

[0059] The model training module is used to train a precise control model for the polymer LPBF process using a database, establish a mapping relationship between process parameters, printing layer thickness and layer-by-layer temperature field distribution, and obtain a negative feedback mechanism between process parameters and real-time temperature field of the printing layer.

[0060] Furthermore, the present invention employs the above-mentioned control system to improve the accuracy of polymer laser powder bed melting forming, comprising: building a precise control model of the polymer LPBF process on the LPBF equipment, and improving the heat accumulation effect in the polymer LPBF process by real-time control of printing process parameters, reducing the warpage deformation of polymer materials during the forming process, and improving the accuracy of the formed parts.

[0061] Specifically, see Figure 2 The control strategy adopted in this invention is as follows:

[0062] Based on the difference between the temperature distribution of each pixel on the single-track laser scanning path and the target temperature distribution during the printing process, the laser energy density and lamp power are adjusted in real time. The laser energy density is jointly controlled by the laser power, scanning speed and scanning spacing.

[0063] Preferably, the algorithm used is a reinforcement learning algorithm; the language used is Python; and the software used is PyCharm.

[0064] Preferably, the polymer material can be one or a mixture of several of the following: polyaryletherketone (PEAK), polyetherketone (PEK), polyetherketoneketone (PEKK), polyetheretherketoneketone (PEEKK), polyimide (PI), polycarbonate (PC), polyethylene (PE), polypropylene (PP), polystyrene (PS), and nylon (PA).

[0065] Preferably, the database consists of multiple episodes, where each episode is defined as the entire process of completing one laser single-channel scan.

[0066] Preferably, the environmental state function is ,in, x , y Used to determine the position of each pixel in the printed layer of the formed part during the printing process. z For the current printed layer thickness, This indicates the temperature distribution of the printed layer on the molded part.

[0067] Preferably, the action range A of the model is defined as a set of process parameters that can be adjusted in real time, including lamp power, laser power, scanning speed, and scanning spacing.

[0068] Preferably, the environmental reward R is defined as the difference between the temperature distribution state of each pixel in the next layer and the target temperature distribution state after a single-channel laser scan is completed.

[0069] Preferably, the value function is ,in, This represents the strategy executed by an individual, specifically the temperature distribution state of the printed layer on the molded part. s At that time, a certain controllable process parameter within the action range A is adjusted to... a The probability is used to explore better parameters; Refer to the execution of operations a Temperature distribution of the next layer of the post-formed part; In the next state The process parameters that need to be adjusted; It is to perform operations a Later in state s The rewards observed in the process; This is a decay factor used to control the weight of the current reward and the rewards for subsequent states; It represents the expected value of the value function, used to measure the magnitude of the theoretical reward and to provide a basis for the model to choose subsequent actions.

[0070] Preferably, taking advantage of the reward-maximizing property of the value function, the absolute error between the temperature distribution on the single-channel laser scanning path and the target temperature distribution is summed as a decreasing term of the value function.

[0071] Preferably, a regularization term is designed into the value function to reduce the variance in the model error and solve the overfitting problem.

[0072] The following embodiment of the invention uses 10% h-BN / PEEK composite powder as an example to conduct related molding accuracy control experiments. Its melting point is 340 ℃. During the LPBF process, it is significantly affected by the heat accumulation effect, and the accuracy of the molded component is significantly affected. The specific preparation method includes the following steps:

[0073] PEEK powder was dried at 120°C for 12 hours using a dryer to ensure it was completely dry. A 120-mesh sieve was then used to evenly pour the PEEK powder through it. The sieve was then placed on a vibrating screen for 300 seconds and removed to obtain fine PEEK powder. The same process was repeated to obtain fine h-BN powder. 2700g of PEEK powder and 300g of h-BN powder were weighed separately, preliminarily mixed, and then poured into a powder mixer. After mixing for 3 hours, the mixture was removed to obtain 10% h-BN / PEEK composite powder. The prepared powder should be sealed and stored in a dry, dark environment to prevent moisture absorption or contamination.

[0074] Example 1

[0075] 1) A standard LPBF (Liquid-to-Balance) forming experiment was conducted using 10% h-BN / PEEK composite powder. Temperature distribution data during the process was acquired using an infrared thermal imager. Process parameters included preheating temperature, laser power, scanning speed, scanning interval, and lamp power. Specifically, the preheating temperatures were 300℃, 310℃, and 320℃; the laser powers were 3 W, 6 W, 9 W, 12 W, 15 W, 21 W, 24 W, 27 W, and 30 W; the scanning speeds were 2000 mm / s, 2500 mm / s, 3000 mm / s, and 3500 mm / s; the scanning intervals were 0.1 mm, 0.2 mm, and 0.3 mm; and the lamp power was adjusted based on the intensification of heat accumulation during the experiment (see [reference needed]). Figure 4 (This is a scatter plot showing the temperature change inside the forming cavity during the printing process). When the formed part does not warp and the surrounding powder does not clump (see...), Figure 5 (This is a topographic image of the formed part before the bottom heat source is formed in the initial stage of printing). The temperature distribution state corresponding to the single laser scanning path of the printed layer of the formed part is extracted as a set of data and combined with the current corresponding process parameters to form a set of episodes (see [link]). Figure 6 (This involves extracting the schematic diagram for each episode in the database). As printing progresses, defects such as warping and agglomeration of the formed part and surrounding powder may occur (see...). Figure 7(To obtain the topographic image of the formed part after printing 10 layers to form the bottom heat source), extract the data again to form a set of episodes, repeat the above standard experiment, and complete the establishment of the database.

[0076] 2) Based on machine learning, establish a precise control model for polymer LPBF process.

[0077] The environment state function is ,in, x , y Used to determine the position of each pixel in the printed layer of the formed part during the printing process. z For the current printed layer thickness, This indicates the temperature distribution of the printed layer on the molded part.

[0078] The model's action range A is defined as a set of process parameters that can be adjusted in real time, including lamp power, laser power, scanning speed, and scanning spacing.

[0079] The environmental reward R is defined as the difference between the temperature distribution state of each pixel in the next layer and the target temperature distribution state after a single laser scan is completed.

[0080] The value function is ,in, This represents the strategy executed by an individual, specifically the temperature distribution state of the printed layer on the molded part. s At that time, a certain controllable process parameter within the action range A is adjusted to... a The probability is used to explore better parameters; Refer to the execution of operations a Temperature distribution of the next layer of the post-formed part; In the next state The process parameters that need to be adjusted; It is to perform operations a Later in state s The rewards observed in the process; This is a decay factor used to control the weight of the current reward and the rewards for subsequent states; This represents the expected value of the value function, used to measure the magnitude of the theoretical reward and providing a basis for the model to select subsequent actions. Utilizing the reward-maximizing property of the value function, the absolute errors between the temperature distribution along a single laser scanning path and the target temperature distribution are summed as a decreasing term in the value function.

[0081] 3) Use the database to train a precise control model for the polymer LPBF process, establish the mapping relationship between process parameters, printing layer thickness and layer-by-layer temperature field distribution, and construct a negative feedback mechanism of process parameters-real-time temperature field.

[0082] 4) A precise control model for the polymer LPBF process is built on the LPBF equipment. Based on the real-time temperature distribution of the printed layer read by the infrared thermal imager after the first layer is printed, the model appropriately adjusts the process parameters for the next layer, controlling the temperature on the single-pass laser scanning path within the optimal temperature window to improve the heat accumulation effect in the polymer LPBF process and ultimately improve the accuracy of the formed parts. Figure 8 ).

[0083] Comparative Example 1

[0084] The printing temperature range for the 10% h-BN / PEEK composite powder should be set between the initial melting temperature and the initial crystallization temperature, i.e., between 302 and 320 °C. Therefore, a 1 mm thick layer of powder should be pre-laid inside the forming cylinder, and the internal temperature of the forming cavity should be set to 300 °C for one hour. Subsequently, the laser power is set to 18 W, the scanning speed to 3000 mm / s, and the scanning interval to 0.1 mm. The internal temperature of the forming cavity is maintained at 320 °C throughout the printing process. After printing, the internal temperature of the forming cavity is slowly cooled to 300 °C at a rate of 1 °C / min, and then held at that temperature for half an hour before the formed part is removed to improve the accuracy of the formed part. Figure 9 ).

[0085] Comparing the molded parts obtained by printing the same material using the printing methods in Example 1 and Comparative Example 1, it can be found that the precision of the molded part obtained by the method of the present invention in Example 1 is significantly better than that of the molded part obtained by direct printing in Comparative Example 1. This is because Comparative Example 1 does not consider the influence of heat accumulation effect and the adaptive control of real-time process parameters. As printing proceeds, the heat of the molded part and the surrounding powder continuously increases, and the bottom heat source gradually forms, causing defects such as warping of the molded part and powder agglomeration, which ultimately leads to a decrease in the precision of the molded part.

[0086] In summary, this invention acquires temperature distribution data during the polymer LPBF process; constructs a database based on the temperature distribution data and the process parameters used for printing the current layer; establishes a precise control model for the polymer LPBF process based on machine learning; and trains the precise control model using the database, establishing a mapping relationship between process parameters, printing layer thickness, and layer-by-layer temperature field distribution, thus constructing a negative feedback mechanism between process parameters and the real-time temperature field. Furthermore, the precise control model for the polymer LPBF process constructed using the above control method is applied to an LPBF device to achieve adaptive control of polymer LPBF forming process parameters. This invention can monitor and precisely control the temperature field distribution during the polymer LPBF process in real time. Utilizing the superiority of reinforcement learning algorithms in big data processing, it improves the heat accumulation phenomenon during the forming process, eliminates defects such as warping and agglomeration, and has guiding significance for improving the accuracy and performance of polymer LPBF formed parts.

[0087] The above content is only for illustrating the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solution based on the technical concept proposed in this invention shall fall within the scope of protection of the claims of this invention.

Claims

1. A method of controlling the precision of polymer laser powder bed fusion, characterized in that, The method comprises the following steps: S1: acquiring temperature distribution data in the polymer LPBF process; S2: constructing a database based on the temperature distribution data and the process parameters used to print the current layer; S3: based on machine learning, the machine learning adopts a reinforcement learning algorithm, the database is composed of multiple episodes, wherein each episode is defined as the entire process of completing a single laser scanning; an accurate control model of the polymer LPBF process is established, including an environmental state function S, an action range A of the model, a reward R of the environment, and a value function Q; wherein the environmental state function is wherein, x , y for determining the position of each pixel point of the real-time printing layer of the formed part in the printing process, z is the current printing layer thickness, represents the temperature distribution state of the real-time printing layer of the formed part; The action range A of the model is defined as a set of process parameters that can be real-time controlled, including lamp power, laser power, scanning speed and scanning pitch; The reward R of the environment is the difference between the temperature distribution state of each pixel point of the next layer and the target temperature distribution state after completing a single laser scanning; The value function Q is wherein: represents the strategy executed by the individual, i.e. the probability of adjusting a controllable process parameter a in the action range A when the temperature distribution state of the printed layer of the shaped part is s , for exploring better parameters; represents the executed operation a , the temperature distribution state of the next layer of the shaped part; is the process parameter that needs to be regulated in the next state ; is the reward observed in the state a after the executed operation s ; is a decay factor for controlling the weight of the current reward and the reward of the subsequent state; represents the expected value of the value function, for measuring the size of the theoretical reward, giving the model the basis for selecting the subsequent action; S4: using the database to train the accurate control model of the polymer LPBF process, establishing the mapping relationship between the process parameters, the printing layer thickness and the temperature field distribution of each layer, obtaining the negative feedback mechanism of the process parameters and the real-time temperature field of the printing layer, and completing the accurate control of the polymer laser powder bed fusion forming precision.

2. The method of claim 1, wherein the polymer laser powder bed fusion precision is controlled by: In S1, the temperature distribution data includes the printing layer thickness and the temperature distribution state of each pixel point of the real-time printing layer of the formed part; According to the difference between the temperature distribution state of each pixel point on the single laser scanning path in the printing process and the target temperature distribution state, the laser energy density and the lamp power are real-time controlled, wherein the laser energy density is controlled by the laser power, the scanning speed and the scanning pitch.

3. The method of claim 1, wherein the method is performed in a powder bed fusion system. In S2, the process parameters include preheating temperature, lamp power, laser power, scanning speed and scanning pitch.

4. The method of claim 1, wherein the method is performed in a polymer laser powder bed fusion process. The absolute error between the temperature distribution on the single laser scanning path and the target temperature distribution is summed as the decreasing term of the value function by maximizing the reward characteristics of the value function.

5. The method of claim 1, wherein the method is performed in a polymer laser powder bed fusion precision control system. In S4, the negative feedback mechanism of the process parameters and the real-time temperature field of the printing layer.

6. The method of claim 1-5, wherein the method is characterized by, The polymer includes one or more of polyaryletherketone, polyetherketone, polyetherketone ketone, polyether ether ketone ketone, polyimide, polycarbonate, polyethylene, polypropylene, polystyrene and nylon.

7. A control system based on the control method of the precision of polymer laser powder bed fusion according to any one of claims 1 to 6, characterized in that, It comprises: a data acquisition module for acquiring temperature distribution data in the polymer LPBF process; a database construction module for constructing a database based on the temperature distribution data and the process parameters used to print the current layer; a control model establishment module for establishing an accurate control model of the polymer LPBF process based on machine learning; a model training module for training the accurate control model of the polymer LPBF process using the database, establishing the mapping relationship between the process parameters, the printing layer thickness and the temperature field distribution of each layer, and obtaining the negative feedback mechanism of the process parameters and the real-time temperature field of the printing layer.

8. The method of claim 7 for improving the precision of polymer laser powder bed fusion forming using the regulation system, characterized in that, It comprises: The accurate control model of the polymer LPBF process is built on the LPBF equipment, the printing process parameters are real-time controlled, the thermal accumulation effect in the polymer LPBF process is improved, the warping deformation of the polymer material in the forming process is reduced, and the precision of the formed part is improved.

Images