Die-free mirror-image-based fiber placement method and die-free mirror-image-based fiber placement system

Abstract

The present invention belongs to the technical field of material machining, and relates to a die-free mirror-image-based fiber placement method and a die-free mirror-image-based fiber placement system, which place fibers in a mirror symmetry manner without requiring dies, thereby improving the flexibility and efficiency of production. The method comprises: acquiring in real time the temperature and pressure of fibers; by using a temperature difference and a pressure difference, calculating the output power of a heater and a modified value of a path of a fiber placement tail end movement mechanism; and optimizing controller parameters by means of reinforcement learning, so as to realize stable control over the temperature and pressure. The system comprises a hardware portion and a control system, wherein the hardware portion includes components such as a pressure component and a heat source, and the control system performs adjustment on the basis of real-time feedback, thereby ensuring the quality of fiber placement. The present invention solves the problem of it being difficult to accurately control the temperature and pressure during die-free mirror-image-based fiber placement, and improves the performance of composite products.

Description

A moldless mirror-image fiber placement method and a moldless mirror-image fiber placement system Technical Field

[0001] This invention belongs to the field of materials processing technology, and relates to a moldless mirror filament placement method and a moldless mirror filament placement system. Background Technology

[0002] In the field of materials processing, wire layup technology, as a key process in composite material manufacturing, is widely used in the production of complex components such as aircraft wings. Traditional wire layup technology relies on pre-manufactured metal molds, which need to be customized according to the curved shape of the component. This is not only time-consuming and material-intensive, but also requires an independent mold for each different component, which greatly limits the flexibility and efficiency of production. In addition, the design and manufacturing process of the molds is complex and technically demanding, further increasing the technical threshold and cost.

[0003] In the field of materials processing, mirror-image machining technology typically refers to using two or more identical or symmetrical machining equipment to simultaneously perform machining operations on both sides of the workpiece, thereby achieving precise and symmetrical machining. Theoretically, if mirror-image machining technology could be combined with wire placement technology, using two identical wire placement machines located on either side of the workpiece in mirror-symmetrical positions, the two machines could provide mutual support during the wire placement process, eliminating the need for molds and achieving moldless wire placement. However, this idea is extremely difficult to realize because:

[0004] 1. During the fiber placement process, the curing heat is provided by heaters on both sides, resulting in a complex and unstable heat transfer path. As the number of fiber layers increases and the structure of the intermediate layer changes, the heat transfer effect on the opposite side becomes unpredictable, leading to temperature fluctuations. These temperature fluctuations directly affect the curing quality of thermoplastic resins and may cause serious problems such as material performance degradation, brittle cracks, or even combustion.

[0005] 2. If two driven rollers are used to squeeze each other to generate compaction pressure, the pressure is prone to drastic fluctuations due to the power source, small contact area and high curvature of the rollers. Unstable pressure not only affects the filling effect of the resin on the fiber gaps, which may form bubbles or voids and reduce the mechanical properties of the composite material, but may also cause excessive resin extrusion, resulting in uneven interlayer bonding between fibers and weakening the interlayer peel strength of the composite material.

[0006] 3. During the wire laying process, robotic arms, gantry cranes, and other motion mechanisms are usually required. The motion accuracy and rigidity of these motion mechanisms are relatively low, far lower than those of machine tools, which increases the difficulty of pressure and temperature control during the wire laying process.

[0007] Therefore, if we want to combine mirror processing technology with the wire placement process, the problem that needs to be solved is: how to accurately control temperature and pressure during the wire placement process.

[0008] Currently, there are some technologies for controlling temperature and pressure. For example, patent application US202117401138A discloses an in-situ monitoring method for compaction rollers in composite material fiber laying process. It uses an infrared thermal imager to collect temperature data for online monitoring and feedback control. However, this patent application is for traditional die-laid or planar fiber laying scenarios. It infers the defects of composite materials by analyzing the temperature profile of the pressure roller. This method is easily affected by various factors and is not suitable for dieless fiber laying.

[0009] Furthermore, existing pressure control technologies are usually based on molds to predict compaction force (e.g., the literature "Modeling and experimental validation of compaction pressure distribution for automated fiber placement, Composite Structures, Volume 256, 2021, 113101, ISSN 0263-8223" and "Pressure distribution for automated fiber placement and design optimization of compaction rollers. Journal of Reinforced Plastics and Composites. 2019; 38(18): 860-870"), which are also not applicable to moldless fiber placement. Summary of the Invention

[0010] The purpose of this invention is to solve the problems existing in the prior art and to provide a moldless mirror filament placement method and a moldless mirror filament placement system.

[0011] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0012] A moldless mirror-image yarn laying method, which lays two ribbons in a mirror-symmetrical manner without using a mold, and repeats the yarn laying process multiple times;

[0013] The actual ribbon temperature is acquired in real time during the yarn laying process, and the temperature difference is calculated in real time. Based on the temperature difference, the modified value of the heater output power is calculated in real time, and the heater output power is updated in real time.

[0014] Temperature difference = the difference between the actual ribbon temperature and the target ribbon temperature;

[0015] The modified value of the heater output power = kp*×T + ki*×ΣT + kd*×ΔT;

[0016] In the formula, T is the temperature difference at the current moment of the current filament placement process, ΣT is the sum of all temperature differences up to the current moment of the current filament placement process, and ΔT = temperature difference at the current moment of the current filament placement process - temperature difference at the previous moment of the current filament placement process; kp*, ki*, and kd* are the controller parameter combinations that meet the requirements, which are found by iterative optimization using a reinforcement learning model through simulation in a virtual environment.

[0017] During the yarn laying process, the actual yarn pressure is acquired in real time and the pressure difference is calculated in real time. Based on the pressure difference, the modification value of the path of the yarn laying end motion mechanism (robotic arm or other motion mechanism, such as gantry or any single-axis or multi-axis motion actuator) is calculated in real time, and the path of the yarn laying end motion mechanism is updated in real time.

[0018] Pressure differential = the difference between the actual ribbon pressure and the target ribbon pressure;

[0019] The modification value of the path of the end-of-wire laying motion mechanism = kp'×P + ki'×ΣP + kd'×ΔP;

[0020] In the formula, P is the pressure difference at the current moment of the current fiber placement process, ΣP is the sum of all pressure differences up to the current moment of the current fiber placement process, and ΔP = pressure difference at the current moment of the current fiber placement process - pressure difference at the previous moment of the current fiber placement process; kp', ki', and kd' are the controller parameter combinations that meet the requirements, which are found by iterative optimization using a reinforcement learning model through simulation in a virtual environment.

[0021] As a preferred technical solution:

[0022] In the moldless mirror filament placement method described above, kp*, ki*, and kd* are obtained through the following process:

[0023] (a) Generate a set of kp, ki, kd;

[0024] (b) Let i = 1;

[0025] (c) A simulation identical to the actual process is being performed in virtual space, and the i-th fiber placement process begins;

[0026] (d) Let j = 1, and let the virtual heater output power equal the actual heater output power in the process;

[0027] (e) Input the virtual heater output power and the corresponding heating time into the prediction model A, and output the predicted ribbon temperature. Calculate the difference between the predicted ribbon temperature and the target ribbon temperature to obtain the j-th temperature difference of the i-th temperature difference sequence.

[0028] Prediction model A is a trained deep learning model. During training, the heater output power and corresponding heating time are used as inputs to the deep learning model, and the ribbon temperature is used as the theoretical output of the deep learning model. The parameters of the deep learning model are continuously adjusted.

[0029] (f) Determine whether the i-th wire laying process is completed. If not, calculate the modified value of the virtual heater output power using the following formula, update the virtual heater output power, and set j = j + 1 before returning to step (e). Otherwise, proceed to the next step. Modified value of virtual heater output power = kp × T + ki × Σ T + kd × Δ T.

[0030] In the formula, T is the j-th temperature difference of the i-th temperature difference sequence, ΣT is the sum of all temperature differences of the i-th temperature difference sequence, and ΔT = j-th temperature difference of the i-th temperature difference sequence - (j-1)-th temperature difference of the i-th temperature difference sequence. When j = 1, let ΔT = 0.

[0031] (g) Score the i-th temperature difference sequence and determine whether the score of the i-th temperature difference sequence reaches the set value. If it does, output the last group of kp, ki, kd and use it as kp*, ki*, kd*; otherwise, proceed to the next step.

[0032] (h) Determine whether all the wire laying processes are completed. If not, input the last set of kp, ki, kd and the i-th temperature difference sequence into the reinforcement learning model with a reward function, and let it output a new set of kp, ki, kd. Let i = i + 1 and return to step (c). Otherwise, report an error, inform the technician that no suitable parameters were found, and return to step (a).

[0033] The process for obtaining prediction model A using the moldless mirror fiber placement method described above is as follows:

[0034] (i) Create a dataset;

[0035] Collect the heater output power, corresponding heating time, and actual ribbon temperature from historical processes that are identical to the current process;

[0036] At the same time, the same simulation as the actual process is performed in the virtual space to obtain the virtual heater output power and corresponding heating time, and the virtual ribbon temperature;

[0037] (ii) Data augmentation: Increase the size of the dataset to obtain a training set;

[0038] (iii) Establish and train a deep learning model to obtain prediction model A.

[0039] In the moldless mirror filament placement method described above, kp', ki', and kd' are obtained through the following process:

[0040] (A) Generate a set of kp, ki, kd;

[0041] (B) Let i = 1;

[0042] (C) A simulation identical to the actual process is being performed in virtual space, and the i-th wire placement process begins;

[0043] (D) Let j = 1, and let the virtual wire placement end motion mechanism path = the actual wire placement end motion mechanism path in the process;

[0044] (E) Input the virtual yarn-laying end motion mechanism path into the prediction model B, and output the predicted ribbon pressure. Calculate the difference between the predicted ribbon pressure and the target ribbon pressure to obtain the j-th pressure difference of the i-th pressure difference sequence.

[0045] Prediction model B is a trained deep learning model. During training, the path of the wire-laying end motion mechanism is used as the input of the deep learning model, and the ribbon pressure is used as the theoretical output of the deep learning model. The parameters of the deep learning model are continuously adjusted.

[0046] (F) Determine whether the i-th filament placement process is completed. If not, calculate the modified value of the virtual filament placement end motion mechanism path using the following formula, update the virtual filament placement end motion mechanism path, and set j = j + 1 before returning to step (E); otherwise, proceed to the next step. Modified value of the virtual filament placement end motion mechanism path = kp × P + ki × Σ P + kd × Δ P;

[0047] In the formula, P is the j-th pressure difference of the i-th pressure difference sequence, ΣP is the sum of all pressure differences of the i-th pressure difference sequence, ΔP = j-th pressure difference of the i-th pressure difference sequence - (j-1)-th pressure difference of the i-th pressure difference sequence, and when j = 1, let ΔP = 0;

[0048] (G) Score the i-th pressure difference sequence and determine whether the score of the i-th pressure difference sequence reaches the set value. If it does, output the last group of kp, ki, kd and use it as kp', ki', kd'; otherwise, proceed to the next step.

[0049] (H) Determine whether all the wire laying processes are completed. If not, input the last set of kp, ki, kd and the i-th pressure difference sequence into the reinforcement learning model with a reward function, and let it output a new set of kp, ki, kd. Let i = i + 1 and return to step (C). Otherwise, report an error, inform the technician that no suitable parameters were found, and return to step (A).

[0050] The process for obtaining prediction model B using the moldless mirror fiber placement method described above is as follows:

[0051] (I) Create a dataset;

[0052] Collect data on the end-of-lay motion mechanism path and actual ribbon pressure from historical processes identical to the current process;

[0053] At the same time, the same simulation as the actual process is performed in the virtual space to obtain the virtual motion path of the yarn laying end mechanism and the virtual ribbon pressure;

[0054] (II) Data augmentation: Increase the size of the dataset to obtain a training set;

[0055] (III) Establish and train a deep learning model to obtain prediction model B.

[0056] The present invention also provides a moldless mirror fiber placement system, including hardware components and a control system;

[0057] The hardware consists of a symmetrical left and right section. The right section includes a main plate, pressure components, shearing components, refeeding components, tension components, heat source components, connectors, and a yarn-laying end motion mechanism.

[0058] Pressure components include pressure sensors;

[0059] The heat source components include a heater and an infrared thermal imager. The heater is used to heat the ribbon, and the infrared thermal imager is used to capture the temperature field data of the ribbon in real time in the area where the ribbon is laid.

[0060] The control system includes a temperature stability control subsystem, a pressure stability control subsystem, a control command generation subsystem, a fiber placement end motion mechanism control subsystem, and a fiber placement end functional mechanism control subsystem.

[0061] Infrared thermal imagers are used to acquire the real temperature of the ribbon in real time during the yarn laying process and send the real temperature of the ribbon to the temperature stabilization control subsystem in real time.

[0062] The temperature stabilization control subsystem is used to calculate the temperature difference in real time, calculate the modified value of the heater output power based on the temperature difference in real time, and send the modified value of the heater output power to the control command generation subsystem in real time.

[0063] Temperature difference = the difference between the actual ribbon temperature and the target ribbon temperature; Modified value of heater output power = kp*×T + ki*×ΣT + kd*×ΔT;

[0064] In the formula, T is the temperature difference at the current moment of the current filament placement process, ΣT is the sum of all temperature differences up to the current moment of the current filament placement process, and ΔT = temperature difference at the current moment of the current filament placement process - temperature difference at the previous moment of the current filament placement process; kp*, ki*, and kd* are the controller parameter combinations that meet the requirements, which are found by iterative optimization using a reinforcement learning model through simulation in a virtual environment.

[0065] The pressure sensor is used to acquire the real ribbon pressure in real time during the yarn laying process and send the real ribbon pressure to the pressure stabilization control subsystem in real time.

[0066] The pressure stabilization control subsystem is used to calculate the pressure difference in real time, calculate the modification value of the path of the yarn placement end motion mechanism in real time based on the pressure difference, and send the modification value of the path of the yarn placement end motion mechanism to the control command generation subsystem in real time.

[0067] Pressure difference = the difference between the actual ribbon pressure and the target ribbon pressure; Modification value of the path of the yarn laying end motion mechanism = kp'×P + ki'×ΣP + kd'×ΔP;

[0068] In the formula, P is the pressure difference at the current moment of the current fiber placement process, ΣP is the sum of all pressure differences up to the current moment of the current fiber placement process, ΔP = pressure difference at the current moment of the current fiber placement process - pressure difference at the previous moment of the current fiber placement process; kp', ki', and kd' are the controller parameter combinations that meet the requirements, which are found by iterative optimization using a reinforcement learning model through simulation in a virtual environment.

[0069] The control command generation subsystem generates control commands based on the modified values ​​of the heater output power and the modified values ​​of the path of the filament placement end motion mechanism, and then sends them to the filament placement end functional mechanism control subsystem and the filament placement end motion mechanism control subsystem.

[0070] The control subsystem for the end-of-wire laying function is used to process control commands and update the heater output power in real time.

[0071] The control subsystem for the end-of-wire placement motion mechanism is used to process control commands and update the path of the end-of-wire placement motion mechanism in real time.

[0072] As a preferred technical solution:

[0073] As described above, the moldless mirror fiber placement system includes a pressure roller, a support device, a heat-insulating asbestos board, and a cylinder a. The central axis of the pressure roller is parallel to the front-to-back direction. The pressure roller, support device, heat-insulating asbestos board, pressure sensor, and cylinder a are connected sequentially from left to right. Cylinder a is fixed to the main body plate and is used to push the pressure roller to the left, cooperating with the pressure roller on the opposite side to squeeze and generate compaction pressure. The heat-insulating asbestos board can isolate the heat on the pressure roller to prevent damage to the pressure sensor.

[0074] The cutting component includes a cylinder C and a blade. The cylinder C is used to drive the blade to cut the ribbon. The cylinder C is fixed to the main body plate.

[0075] The feeding component includes a drive wheel, a driven wheel, a motor, and cylinder b. Both the drive wheel and driven wheel are vertically arranged and parallel to the left-right direction. The motor drives the drive wheel to rotate, and cylinder b presses the driven wheel onto the drive wheel. The motor and cylinder b are fixed to the main body plate. Before the ribbon is installed and the yarn is laid, cylinder b is not activated, and the drive wheel and driven wheel are separated. When the yarn laying begins, the ribbon is installed in the channel and passes between the drive wheel and driven wheel. Then, cylinder b is activated, and the driven wheel presses the ribbon onto the drive wheel, achieving ribbon compression. The motor drives the drive wheel to rotate during the yarn laying process, causing the driven wheel to extrude the ribbon together.

[0076] The tensioning components include a material roll and a brake. The material roll is used to wind the ribbon. During the yarn laying process, the material roll is forced to rotate under the drive of the heavy feed component to feed the ribbon. The brake is used to provide torque to the material roll to maintain a stable tension force for the yarn feeding, so as to prevent the ribbon from being unstablely fed or even falling off due to the high speed rotation of the material roll. Both the material roll and the brake are fixed on the main body plate.

[0077] The heat source components also include a support plate, a heater fixed to the main body plate, and an infrared thermal imager fixed to the main body plate via the support plate;

[0078] The connector is fixed to the main plate and connected to the end of the wire laying end motion mechanism. Beneficial effects:

[0079] (1) This invention eliminates the reliance on molds in traditional wire placement technology through moldless mirror wire placement technology, avoids the time and material costs required for mold manufacturing, and avoids the problem that different components require different molds. This makes production more flexible and can quickly adapt to diverse component shapes, thereby significantly improving production efficiency.

[0080] (2) This invention obtains real-time temperature and pressure of the ribbon and optimizes the controller parameter combination obtained by reinforcement learning model, calculates and adjusts the output power of the heater and the path of the motion mechanism at the end of the filament laying in real time, ensuring stable control of temperature and pressure during the moldless mirror filament laying process. This effectively avoids problems such as uneven or too fast curing and material performance degradation caused by temperature fluctuations, as well as defects such as poor bonding performance and air bubbles caused by pressure fluctuations, thereby significantly improving the final performance of composite material products. Attached Figure Description

[0081] Figure 1 is a front view of the hardware structure of the moldless mirror filament placement system of the present invention;

[0082] Figure 2 is a rear view of the left part of the hardware component of the moldless mirror filament placement system of the present invention.

[0083] Figure 3 is a front view of the right part of the hardware component of the moldless mirror filament placement system of the present invention.

[0084] Figure 4 is a schematic diagram of the ribbon installation process;

[0085] Figure 5 shows a comparison of the fiber placement distance-temperature curves between experimental group 1 and control group 1;

[0086] Figure 6 shows a comparison of the fiber placement distance-temperature curves between experimental group 2 and control group 2;

[0087] Figure 7 shows a comparison of the fiber placement distance-temperature curves between experimental group 3 and control group 3;

[0088] Figure 8 shows a comparison of the fiber placement distance-pressure curves between experimental group 4 and control group 4;

[0089] Figure 9 shows a comparison of the fiber placement distance-pressure curves between experimental group 5 and control group 5;

[0090] Figure 10 shows a comparison of the fiber placement distance-pressure curves between experimental group 6 and control group 6;

[0091] Figure 11 is a complete flowchart of the present invention. The loop from generating the system control instruction file to determining whether the wire laying is completed is the wire laying process flow, which includes reward calculation, model training, PID parameter (i.e., kp, ki, kd below) adjustment, and PID parameter decision-making process.

[0092] Figure 12 is a logic diagram of the reward calculation, model training, and PID parameter adjustment of the present invention;

[0093] Among them, 1-pressure sensor, 2-cylinder a, 3-connector, 4-material roll, 5-support plate, 6-heater, 7-infrared thermal imager, 8-cylinder c, 9-pressure roller, 10-drive wheel, 11-cylinder b, 12-driven wheel, 13-motor, 14-blade, 15-brake, 16-heat insulation asbestos board, 17-main body plate, 19-ribbon, 20-bracket. Detailed Implementation

[0094] The present invention will be further described below with reference to specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. Furthermore, it should be understood that after reading the teachings of this invention, those skilled in the art can make various alterations or modifications to the invention, and these equivalent forms also fall within the scope defined by the appended claims.

[0095] A moldless mirror filament placement system, as shown in Figures 1, 2, 3, 11 and 12, includes a hardware component and a control system.

[0096] As shown in Figures 1 to 3, the hardware part includes a symmetrical left part and a right part. The right part includes a main plate 17, a pressure component, a shearing component, a refeeding component, a tension component, a heat source component, a connector 3, and a filament laying end motion mechanism.

[0097] As shown in Figures 1 and 3, the pressure component includes a pressure sensor 1, a pressure roller 9, a support device, a heat-insulating asbestos board 16, and a cylinder a 2; the central axis of the pressure roller 9 is parallel to the front-back direction, and the pressure roller 9, the support device, the heat-insulating asbestos board 16, the pressure sensor 1, and the cylinder a 2 are connected in sequence from left to right. The cylinder a 2 is fixed on the main body plate 17 and is used to push the pressure roller 9 to the left.

[0098] The cutting component includes a cylinder c8 and a blade 14. The cylinder c8 is used to drive the blade 14 to cut the ribbon 19. The cylinder c8 is fixed on the main body plate 17.

[0099] As shown in Figures 2 and 3, the heavy conveying component includes a drive wheel 10, a driven wheel 12, a motor 13, and a cylinder b 11. The drive wheel 10 and the driven wheel 12 are both arranged vertically and parallel to the left and right directions. The motor 13 is used to drive the drive wheel 10 to rotate, and the cylinder b 11 is used to press the driven wheel 12 onto the drive wheel 10. The motor 13 and the cylinder b 11 are fixed on the main body plate 17.

[0100] The tensioning components include a material roll 4 and a brake 15. The material roll 4 is used to wind the ribbon 19, and the brake 15 is used to provide torque to the material roll 4 to maintain a stable tension force for the wire feeding. Both the material roll 4 and the brake 15 are fixed on the main body plate 17.

[0101] The heat source components include a heater 6, an infrared thermal imager 7, and a support plate 5; the heater 6 is used to heat the ribbon 19; the infrared thermal imager 7 is used to capture real-time temperature field data of the ribbon 19 in the yarn-laying area; the heater 6 is fixed on the main body plate 17, and the infrared thermal imager 7 is fixed on the main body plate 17 through the support plate 5.

[0102] Connector 3 is fixed on the main plate 17 and connected to the end of the filament laying end motion mechanism;

[0103] The control system includes a temperature stability control subsystem, a pressure stability control subsystem, a control command generation subsystem, a fiber placement end motion mechanism control subsystem, and a fiber placement end functional mechanism control subsystem.

[0104] The infrared thermal imager 7 is also used to acquire the real ribbon temperature in real time during the ribbon laying process and send the real ribbon temperature to the temperature stabilization control subsystem in real time. The temperature stabilization control subsystem is used to calculate the temperature difference in real time, calculate the modified value of the heater output power based on the temperature difference, and send the modified value of the heater output power to the control command generation subsystem in real time. Temperature difference = the difference between the real ribbon temperature and the target ribbon temperature; Modified value of heater output power = kp*×T + ki*×ΣT + kd*×ΔT;

[0105] In the formula, T is the temperature difference at the current moment of the current fiber placement process, ΣT is the sum of all temperature differences up to the current moment of the current fiber placement process, and ΔT = temperature difference at the current moment of the current fiber placement process - temperature difference at the previous moment of the current fiber placement process; kp*, ki*, and kd* are obtained through the following process:

[0106] (a) Generate a set of kp, ki, kd;

[0107] (b) Let i = 1;

[0108] (c) A simulation identical to the actual process is being performed in virtual space, and the i-th fiber placement process begins;

[0109] (d) Let j = 1, and let the virtual heater output power equal the actual heater output power in the process;

[0110] (e) Input the virtual heater output power and the corresponding heating time into the prediction model A, and output the predicted ribbon temperature. Calculate the difference between the predicted ribbon temperature and the target ribbon temperature to obtain the j-th temperature difference of the i-th temperature difference sequence.

[0111] The process for obtaining prediction model A is as follows:

[0112] (i) Create a dataset;

[0113] Collect the heater output power, corresponding heating time, and actual ribbon temperature from historical processes that are identical to the current process;

[0114] At the same time, the same simulation as the actual process is performed in the virtual space to obtain the virtual heater output power and corresponding heating time, and the virtual ribbon temperature;

[0115] (ii) Data augmentation: Increase the size of the dataset to obtain a training set;

[0116] (iii) Establish and train a deep learning model to obtain prediction model A;

[0117] (f) Determine whether the i-th wire laying process is completed. If not, calculate the modified value of the virtual heater output power using the following formula, update the virtual heater output power, and set j = j + 1 before returning to step (e). Otherwise, proceed to the next step. Modified value of virtual heater output power = kp × T + ki × Σ T + kd × Δ T.

[0118] In the formula, T is the j-th temperature difference of the i-th temperature difference sequence, ΣT is the sum of all temperature differences of the i-th temperature difference sequence, and ΔT = j-th temperature difference of the i-th temperature difference sequence - (j-1)-th temperature difference of the i-th temperature difference sequence. When j = 1, let ΔT = 0.

[0119] (g) Score the i-th temperature difference sequence and determine whether the score of the i-th temperature difference sequence reaches the set value. If it does, output the last group of kp, ki, kd and use it as kp*, ki*, kd*; otherwise, proceed to the next step.

[0120] (h) Determine whether all the wire laying processes are completed. If not, input the last set of kp, ki, kd and the i-th temperature difference sequence into the reinforcement learning model with a reward function, and let it output a new set of kp, ki, kd. Let i = i + 1 and return to step (c); otherwise, return to step (a).

[0121] Pressure sensor 1 is used to acquire the real ribbon pressure in real time during the yarn laying process and send the real ribbon pressure to the pressure stabilization control subsystem in real time; the pressure stabilization control subsystem is used to calculate the pressure difference in real time, calculate the modification value of the yarn laying end motion mechanism path in real time based on the pressure difference, and send the modification value of the yarn laying end motion mechanism path to the control command generation subsystem in real time.

[0122] Pressure differential = the difference between the actual ribbon pressure and the target ribbon pressure;

[0123] The modification value of the path of the end-of-wire laying motion mechanism = kp'×P + ki'×ΣP + kd'×ΔP;

[0124] In the formula, P is the pressure difference at the current moment of the current fiber placement process, ΣP is the sum of all pressure differences up to the current moment of the current fiber placement process, and ΔP = pressure difference at the current moment of the current fiber placement process - pressure difference at the previous moment of the current fiber placement process; kp', ki', and kd' are obtained through the following process:

[0125] (A) Generate a set of kp, ki, kd;

[0126] (B) Let i = 1;

[0127] (C) A simulation identical to the actual process is being performed in virtual space, and the i-th wire placement process begins;

[0128] (D) Let j = 1, and let the virtual wire placement end motion mechanism path = the actual wire placement end motion mechanism path in the process;

[0129] (E) Input the virtual yarn-laying end motion mechanism path into the prediction model B, and output the predicted ribbon pressure. Calculate the difference between the predicted ribbon pressure and the target ribbon pressure to obtain the j-th pressure difference of the i-th pressure difference sequence.

[0130] The process for obtaining prediction model B is as follows:

[0131] (I) Create a dataset;

[0132] Collect data on the end-of-lay motion mechanism path and actual ribbon pressure from historical processes identical to the current process;

[0133] At the same time, the same simulation as the actual process is performed in the virtual space to obtain the virtual motion path of the yarn laying end mechanism and the virtual ribbon pressure;

[0134] (II) Data augmentation: Increase the size of the dataset to obtain a training set;

[0135] (III) Establish and train a deep learning model to obtain prediction model B;

[0136] (F) Determine whether the i-th filament placement process is completed. If not, calculate the modified value of the virtual filament placement end motion mechanism path using the following formula, update the virtual filament placement end motion mechanism path, and set j = j + 1 before returning to step (E); otherwise, proceed to the next step. Modified value of the virtual filament placement end motion mechanism path = kp × P + ki × Σ P + kd × Δ P;

[0137] In the formula, P is the j-th pressure difference of the i-th pressure difference sequence, ΣP is the sum of all pressure differences of the i-th pressure difference sequence, ΔP = j-th pressure difference of the i-th pressure difference sequence - (j-1)-th pressure difference of the i-th pressure difference sequence, and when j = 1, let ΔP = 0;

[0138] (G) Score the i-th pressure difference sequence and determine whether the score of the i-th pressure difference sequence reaches the set value. If it does, output the last group of kp, ki, kd and use it as kp', ki', kd'; otherwise, proceed to the next step.

[0139] (H) Determine whether all the wire laying processes are completed. If not, input the last set of kp, ki, kd and the i-th pressure difference sequence into the reinforcement learning model with a reward function, and let it output a new set of kp, ki, kd. Let i = i + 1 and return to step (C); otherwise, return to step (A).

[0140] The control command generation subsystem generates control commands based on the modified values ​​of the heater output power and the modified values ​​of the path of the filament placement end motion mechanism, and then sends them to the filament placement end functional mechanism control subsystem and the filament placement end motion mechanism control subsystem.

[0141] The control subsystem for the end-of-wire laying function is used to process control commands and update the heater output power in real time.

[0142] The control subsystem for the end-of-wire placement motion mechanism is used to process control commands and update the path of the end-of-wire placement motion mechanism in real time.

[0143] The following describes the steps of using the system of the present invention for fiber placement, taking a process involving two steps as an example:

[0144] (1) Ribbon installation;

[0145] As shown in Figure 4, after the ribbon is wound around the material roll, the ribbon passes between the driving wheel and the driven wheel, passes through the blade and reaches the pressure roller, and is fixed on the bracket 20. The bracket 20 is a vertical frame or a truss fixed on the ground. Then, the connector is connected to the end of the yarn laying end motion mechanism.

[0146] When the ribbon passes between the driven wheel and the driving wheel, cylinder b is activated, pressing the ribbon tightly between the driven wheel and the driving wheel;

[0147] When the ribbon reaches the pressure roller, cylinder a is activated, causing the pressure rollers on both sides to push out and cooperate with the pressure roller on the opposite side to squeeze and tighten the ribbon;

[0148] (2) First silk laying;

[0149] First, the yarn-laying end motion mechanism begins to move along the yarn-laying path. Simultaneously, the heater is activated to heat the yarn, an infrared thermal imager begins to capture real-time temperature field data of the yarn-laying area and monitors the yarn temperature, a pressure sensor acquires the yarn pressure in real time, and a motor drives the drive wheel and driven wheel to rotate, propelling the yarn forward. During this process, the temperature stabilization control subsystem and the pressure stabilization control subsystem calculate the modification values ​​of the heater output power and the yarn-laying end motion mechanism path based on the real-time data, respectively, and send them to the corresponding control subsystems (i.e., the yarn-laying end functional mechanism control subsystem or the yarn-laying end motion mechanism control subsystem) through the control command generation subsystem. The heating power and yarn-laying path are adjusted in real time to ensure that the yarn temperature and pressure remain within the target range.

[0150] Then, when the laying is about to be completed, cylinder C is activated to drive the blade to cut the ribbon and cut off the excess part. After cutting, cylinder C is reset.

[0151] Finally, the reserved ribbon between the blade and the pressure roller continues to be laid until the first laying is completed;

[0152] (3) Second silk laying;

[0153] First, the end-of-wire laying mechanism moves to its initial position;

[0154] Then, the yarn laying end motion mechanism begins to move along the yarn laying path. When the yarn laying is about to be completed, cylinder c is activated to drive the blade to cut the yarn and cut off the excess part. After cutting, cylinder c is reset.

[0155] Finally, the reserved ribbon between the blade and the pressure roller continues to be laid until the second laying is completed. Then, the heater, motor, infrared thermal imager, cylinder a, and cylinder b are turned off, and the connection between the connector and the end of the yarn laying end motion mechanism is disconnected.

[0156] To demonstrate that the control system in the moldless mirror fiber placement system of the present invention can accurately control temperature and pressure, the following experiments were conducted for verification:

[0157] Experimental Group 1: The moldless mirrored yarn placement system of the present invention was used, with the target ribbon temperature set at 230℃, the initial speed of the yarn placement end motion mechanism at 10mm / s, the initial output power of the heater at 90W, and the yarn placement length at 300mm.

[0158] Control group 1: Basically the same as experimental group 1, the only difference being that: there is no control system in the moldless mirror filament placement system;

[0159] Experimental Group 2: Using the moldless mirrored yarn placement system of the present invention, the target ribbon temperature was set to 380℃, the initial speed of the yarn placement end motion mechanism was set to 50mm / s, the initial output power of the heater was set to 210W, and the yarn placement length was set to 300mm.

[0160] Control group 2: Basically the same as experimental group 2, the only difference being that: the moldless mirror filament placement system has no control system;

[0161] Experimental Group 3: Using the moldless mirrored yarn placement system of the present invention, the target ribbon temperature was set to 500℃, the initial speed of the yarn placement end motion mechanism was set to 10mm / s, the initial output power of the heater was set to 135W, and the yarn placement length was set to 300mm.

[0162] Control group 3: Basically the same as experimental group 3, the only difference being that: the moldless mirror fiber placement system has no control system;

[0163] Experimental Group 4: The moldless mirrored yarn placement system of the present invention was used, with the target ribbon pressure set to 15N, the vertical spacing of the yarn placement end motion mechanism set to 4mm, and the yarn placement length set to 300mm.

[0164] Control group 4: Basically the same as experimental group 4, the only difference being that: the moldless mirror filament placement system has no control system;

[0165] Experimental Group 5: The moldless mirrored filament placement system of the present invention was used, with the target ribbon pressure set to 30N, the vertical spacing of the filament placement end motion mechanism set to 4mm, and the filament placement length set to 300mm.

[0166] Control group 5: Basically the same as experimental group 5, the only difference being that: the moldless mirror filament placement system has no control system;

[0167] Experimental Group 6: The moldless mirrored yarn placement system of the present invention was used, with the target ribbon pressure set to 45N, the vertical spacing of the yarn placement end motion mechanism set to 4mm, and the yarn placement length set to 300mm.

[0168] Control group 6: Basically the same as experimental group 6, the only difference being that: the moldless mirror filament placement system has no control system;

[0169] The comparison of the yarn laying distance-temperature curves during the yarn laying process between experimental group 1 and control group 1, experimental group 2 and control group 2, and experimental group 3 and control group 3 are shown in Figures 5, 6 and 7, respectively. As can be seen from Figures 5 to 7, compared with control groups 1 to 3, experimental groups 1 to 3 can stably control the actual ribbon temperature within a range close to the target ribbon temperature.

[0170] The comparison of the yarn laying distance-pressure curves between experimental group 4 and control group 4, experimental group 5 and control group 5, and experimental group 6 and control group 6 during the yarn laying process is shown in Figures 8, 9 and 10, respectively. As can be seen from Figures 8 to 10, compared with control groups 4 to 6, experimental groups 4 to 6 can stably control the actual yarn pressure within a range close to the target yarn pressure.

[0171] The above verification process and results are sufficient to prove that the control system in the moldless mirror filament placement system of the present invention can accurately control temperature and pressure.

Claims

1. A dieless mirror image filament laying method, characterized by, Without using a mold, lay two ribbons in a mirror-symmetrical manner, and repeat this ribbon-laying process multiple times; The actual ribbon temperature is acquired in real time during the yarn laying process, and the temperature difference is calculated in real time. Based on the temperature difference, the modified value of the heater output power is calculated in real time, and the heater output power is updated in real time. Temperature difference = the difference between the actual ribbon temperature and the target ribbon temperature; The modified value of the heater output power = kp*×T + ki*×ΣT + kd*×ΔT; In the formula, T is the temperature difference at the current moment of the current yarn laying process, ΣT is the sum of all temperature differences up to the current moment of the current yarn laying process, and ΔT = temperature difference at the current moment of the current yarn laying process - temperature difference at the previous moment of the current yarn laying process; kp*, ki*, and kd* are controller parameter combinations that meet the requirements and are obtained by simulation in a virtual environment and continuous iterative optimization using a reinforcement learning model. The acquisition process is as follows: start virtual yarn laying with initial kp, ki, and kd, use prediction model A based on the virtual heater output power and corresponding heating time to predict the ribbon temperature to obtain the predicted ribbon temperature, calculate the difference between the predicted ribbon temperature and the target ribbon temperature, and perform closed-loop control by dynamically adjusting the virtual heater output power. If the temperature difference meets the target after a single simulation, output the current parameters as kp*, ki*, and kd*; otherwise, input the last set of kp, ki, kd and temperature difference data into the reinforcement learning model to generate new parameters, re-simulate and continuously iterate. During the yarn laying process, the actual ribbon pressure is acquired in real time and the pressure difference is calculated in real time. Based on the pressure difference, the modification value of the motion mechanism path at the end of the yarn laying is calculated in real time, and the motion mechanism path at the end of the yarn laying is updated in real time. Pressure differential = the difference between the actual ribbon pressure and the target ribbon pressure; The modification value of the path of the end-of-wire laying motion mechanism = kp'×P + ki'×ΣP + kd'×ΔP; In the formula, P is the pressure difference at the current moment of the current yarn laying process, ΣP is the sum of all pressure differences up to the current moment of the current yarn laying process, and ΔP = pressure difference at the current moment of the current yarn laying process - pressure difference at the previous moment of the current yarn laying process; kp', ki', and kd' are controller parameter combinations that meet the requirements, obtained by simulation in a virtual environment and iterative optimization using a reinforcement learning model. The acquisition process is as follows: start virtual yarn laying with initial kp, ki, and kd, obtain the predicted ribbon pressure using prediction model B based on the path of the yarn laying end motion mechanism, calculate the difference between the predicted ribbon pressure and the target ribbon pressure, and perform closed-loop control by dynamically adjusting the virtual yarn laying end motion mechanism path. If the pressure difference meets the target after a single simulation, output the current parameters as kp', ki', and kd'; otherwise, input the last set of kp, ki, kd, and pressure difference data into the reinforcement learning model to generate new parameters, re-simulate, and continue iterating.

2. A dieless mirror imaging filament laying method according to claim 1, wherein, kp*, ki*, and kd* are obtained through the following process: (a) Generate a set of kp, ki, kd; (b) Let i = 1; (c) A simulation identical to the actual process is being performed in virtual space, and the i-th fiber placement process begins; (d) Let j = 1, and let the virtual heater output power equal the actual heater output power in the process; (e) Input the virtual heater output power and the corresponding heating time into the prediction model A, and output the predicted ribbon temperature. Calculate the difference between the predicted ribbon temperature and the target ribbon temperature to obtain the j-th temperature difference of the i-th temperature difference sequence. Prediction model A is a trained deep learning model. During training, the heater output power and corresponding heating time are used as inputs to the deep learning model, and the ribbon temperature is used as the theoretical output of the deep learning model. The parameters of the deep learning model are continuously adjusted. (f) Determine whether the i-th wire laying process is completed. If not, calculate the modified value of the virtual heater output power using the following formula, update the virtual heater output power, and set j = j + 1 before returning to step (e); otherwise, proceed to the next step. The modified value of the virtual heater output power = kp×T + ki×ΣT + kd×ΔT; In the formula, T is the j-th temperature difference of the i-th temperature difference sequence, ΣT is the sum of all temperature differences of the i-th temperature difference sequence, and ΔT = j-th temperature difference of the i-th temperature difference sequence - (j-1)-th temperature difference of the i-th temperature difference sequence. When j = 1, let ΔT = 0. (g) Score the i-th temperature difference sequence and determine whether the score of the i-th temperature difference sequence reaches the set value. If it does, output the last group of kp, ki, kd and use it as kp*, ki*, kd*; otherwise, proceed to the next step. (h) Determine whether all the wire laying processes are completed. If not, input the last set of kp, ki, kd and the i-th temperature difference sequence into the reinforcement learning model with a reward function, and let it output a new set of kp, ki, kd. Let i = i + 1 and return to step (c); otherwise, return to step (a).

3. A dieless mirror imaging filament laying method according to claim 2, wherein, The process for obtaining prediction model A is as follows: (i) Create a dataset; Collect the heater output power, corresponding heating time, and actual ribbon temperature from historical processes that are identical to the current process; At the same time, the same simulation as the actual process is performed in the virtual space to obtain the virtual heater output power and corresponding heating time, and the virtual ribbon temperature; (ii) Data augmentation: Increase the size of the dataset to obtain a training set; (iii) Establish and train a deep learning model to obtain prediction model A.

4. The dieless mirror imaging filament laying method of claim 1, wherein, kp', ki', and kd' are obtained through the following process: (A) Generate a set of kp, ki, kd; (B) Let i = 1; (C) A simulation identical to the actual process is being performed in virtual space, and the i-th wire placement process begins; (D) Let j = 1, and let the virtual wire placement end motion mechanism path = the actual wire placement end motion mechanism path in the process; (E) Input the virtual yarn-laying end motion mechanism path into the prediction model B, and output the predicted ribbon pressure. Calculate the difference between the predicted ribbon pressure and the target ribbon pressure to obtain the j-th pressure difference of the i-th pressure difference sequence. Prediction model B is a trained deep learning model. During training, the path of the wire-laying end motion mechanism is used as the input of the deep learning model, and the ribbon pressure is used as the theoretical output of the deep learning model. The parameters of the deep learning model are continuously adjusted. (F) Determine whether the i-th filament placement process is completed. If not, use the following formula to calculate the modification value of the virtual filament placement end motion mechanism path, update the virtual filament placement end motion mechanism path, and set j = j + 1 before returning to step (E); otherwise, proceed to the next step. The modified value of the virtual wire-laying end motion mechanism path = kp×P + ki×ΣP + kd×ΔP; In the formula, P is the j-th pressure difference of the i-th pressure difference sequence, ΣP is the sum of all pressure differences of the i-th pressure difference sequence, and ΔP = the j-th pressure difference of the i-th pressure difference sequence - the (j-1)-th pressure difference of the i-th pressure difference sequence. When j = 1, let ΔP = 0. (G) Score the i-th pressure difference sequence and determine whether the score of the i-th pressure difference sequence reaches the set value. If it does, output the last group of kp, ki, kd and use it as kp', ki', kd'; otherwise, proceed to the next step. (H) Determine whether all the wire laying processes are completed. If not, input the last set of kp, ki, kd and the i-th pressure difference sequence into the reinforcement learning model with a reward function, and let it output a new set of kp, ki, kd. Let i = i + 1 and return to step (C); otherwise, return to step (A).

5. A dieless mirror imaging filament laying method according to claim 4, wherein, The process for obtaining prediction model B is as follows: (I) Create a dataset; Collect data on the end-of-lay motion mechanism path and actual ribbon pressure from historical processes identical to the current process; At the same time, the same simulation as the actual process is performed in the virtual space to obtain the virtual motion path of the yarn laying end mechanism and the virtual ribbon pressure; (II) Data augmentation: Increase the size of the dataset to obtain a training set; (III) Establish and train a deep learning model to obtain prediction model B.

6. A dieless mirror image filament laying system characterized by, Including hardware components and control systems; The hardware part includes a symmetrical left part and a right part. The right part includes a main plate (17), a pressure component, a shearing component, a refeeding component, a tension component, a heat source component, a connector (3), and a yarn laying end motion mechanism. The pressure component, shearing component, refeeding component, and tension component are arranged in sequence from front to back along the yarn feeding direction. The pressure component includes a pressure sensor (1); The heat source components include a heater (6) and an infrared thermal imager (7). The heater (6) is used to heat the ribbon (19), and the infrared thermal imager (7) is used to capture the temperature field data of the ribbon (19) in real time in the area where the ribbon is laid. The control system includes a temperature stability control subsystem, a pressure stability control subsystem, a control command generation subsystem, a fiber placement end motion mechanism control subsystem, and a fiber placement end functional mechanism control subsystem. The infrared thermal imager (7) is used to acquire the real ribbon temperature in real time during the yarn laying process and send the real ribbon temperature to the temperature stabilization control subsystem in real time. The temperature stabilization control subsystem is used to calculate the temperature difference in real time, calculate the modified value of the heater output power based on the temperature difference in real time, and send the modified value of the heater output power to the control command generation subsystem in real time. Temperature difference = the difference between the actual ribbon temperature and the target ribbon temperature; The modified value of the heater output power = kp*×T + ki*×ΣT + kd*×ΔT; In the formula, T is the temperature difference at the current moment of the current yarn laying process, ΣT is the sum of all temperature differences up to the current moment of the current yarn laying process, and ΔT = temperature difference at the current moment of the current yarn laying process - temperature difference at the previous moment of the current yarn laying process; kp*, ki*, and kd* are controller parameter combinations that meet the requirements and are obtained by simulation in a virtual environment and continuous iterative optimization using a reinforcement learning model. The acquisition process is as follows: start virtual yarn laying with initial kp, ki, and kd, use prediction model A based on the virtual heater output power and corresponding heating time to predict the ribbon temperature to obtain the predicted ribbon temperature, calculate the difference between the predicted ribbon temperature and the target ribbon temperature, and perform closed-loop control by dynamically adjusting the virtual heater output power. If the temperature difference meets the target after a single simulation, output the current parameters as kp*, ki*, and kd*; otherwise, input the last set of kp, ki, kd and temperature difference data into the reinforcement learning model to generate new parameters, re-simulate and continuously iterate. The pressure sensor (1) is used to acquire the real ribbon pressure in real time during the yarn laying process and send the real ribbon pressure to the pressure stabilization control subsystem in real time. The pressure stabilization control subsystem is used to calculate the pressure difference in real time, calculate the modification value of the path of the yarn placement end motion mechanism in real time based on the pressure difference, and send the modification value of the path of the yarn placement end motion mechanism to the control command generation subsystem in real time. Pressure differential = the difference between the actual ribbon pressure and the target ribbon pressure; The modification value of the path of the end-of-wire laying motion mechanism = kp'×P + ki'×ΣP + kd'×ΔP; In the formula, P is the pressure difference at the current moment of the current yarn laying process, ΣP is the sum of all pressure differences up to the current moment of the current yarn laying process, and ΔP = pressure difference at the current moment of the current yarn laying process - pressure difference at the previous moment of the current yarn laying process; kp', ki', and kd' are controller parameter combinations that meet the requirements, obtained by simulation in a virtual environment and iterative optimization using a reinforcement learning model. The acquisition process is as follows: start virtual yarn laying with initial kp, ki, and kd, obtain the predicted ribbon pressure using prediction model B based on the path of the yarn laying end motion mechanism, calculate the difference between the predicted ribbon pressure and the target ribbon pressure, and perform closed-loop control by dynamically adjusting the virtual yarn laying end motion mechanism path. If the pressure difference meets the target after a single simulation, output the current parameters as kp', ki', and kd'; otherwise, input the last set of kp, ki, kd and pressure difference data into the reinforcement learning model to generate new parameters, re-simulate, and continue iterating. The control command generation subsystem generates control commands based on the modified values ​​of the heater output power and the modified values ​​of the path of the filament placement end motion mechanism, and then sends them to the filament placement end functional mechanism control subsystem and the filament placement end motion mechanism control subsystem. The control subsystem for the end-of-wire laying function is used to process control commands and update the heater output power in real time. The control subsystem for the end-of-wire placement motion mechanism is used to process control commands and update the path of the end-of-wire placement motion mechanism in real time.

7. A dieless mirror image filament laying system as claimed in claim 6, wherein, The pressure component also includes a pressure roller (9), a support device, a heat-insulating asbestos board (16), and a cylinder a (2). The central axis of the pressure roller (9) is parallel to the front-back direction. The pressure roller (9), the support device, the heat-insulating asbestos board (16), the pressure sensor (1), and the cylinder a (2) are connected in sequence from left to right. The cylinder a (2) is fixed on the main body plate (17) and is used to push the pressure roller (9) to the left. The cutting component includes a cylinder c (8) and a blade (14). The cylinder c (8) is used to drive the blade (14) to cut the ribbon (19). The cylinder c (8) is fixed on the main body plate (17). The conveying component includes a drive wheel (10), a driven wheel (12), a motor (13), and a cylinder b (11). The drive wheel (10) and the driven wheel (12) are both arranged vertically and parallel to the left and right directions. The motor (13) is used to drive the drive wheel (10) to rotate. The cylinder b (11) is used to press the driven wheel (12) onto the drive wheel (10). The motor (13) and the cylinder b (11) are fixed on the main body plate (17). The tensioning components include a roll (4) and a brake (15). The roll (4) is used to wind the ribbon (19), and the brake (15) is used to provide torque to the roll (4) to maintain a stable tension of the wire feed. Both the roll (4) and the brake (15) are fixed on the main body plate (17). The heat source components also include a support plate (5), a heater (6) fixed on the main body plate (17), and an infrared thermal imager (7) fixed on the main body plate (17) via the support plate (5); The connector (3) is fixed on the main plate (17) and connected to the end of the wire laying end motion mechanism.

Images