Multi-pass flexible flanging forming trajectory optimization method

Abstract

A kind of multi-pass flexible flanging forming trajectory optimization method, according to target flanging piece material setting flanging springback compensation criterion, combined with the multi-pass flexible flanging forming strategy determined according to flanging piece feature, the theoretical flanging forming trajectory and total forming time are comprehensively obtained;In actual flanging mechanism forming process, real-time working condition data are collected by vision system and imported into forming trajectory prediction algorithm, the next time forming trajectory is predicted, closed-loop control of flanging mechanism is realized, until the whole flanging forming process is completed.The method is simple and feasible, can reduce the number of multi-pass flanging springback compensation, avoid multi-pass flanging springback difficult compensation and other problems, effectively improve the forming efficiency and forming quality, and has important engineering application value in aerospace, ship and automobile manufacturing engineering fields.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of sheet forming, in particular to a multi-pass flexible flanging forming trajectory optimization method. BACKGROUND

[0002] Multi-variety small-batch large-size opening sheet metal flanging parts are widely used in the fields of aerospace, automobiles and the like to improve the stiffness of the parts or further assemble with other structures. However, the conventional flanging process is formed on a press machine with a die and a punch, and only large-batch parts are produced. The robot flanging forming technology based on trajectory control combines robot control technology and flanging forming technology, and realizes accurate forming of various flanging by moving the forming tool with the robot, which is particularly suitable for flanging forming of multi-variety small-batch parts. The existing forming method is improved in process, and the springback after the flanging forming is still not compensated because the material is severely work-hardened, so in order to further improve the forming quality, the flanging forming trajectory needs to be optimized to realize accurate compensation of springback during the forming process. The existing laser bending forming technology for aluminum alloy sheets cannot immediately adjust the process parameters, but only modifies the process parameters after the forming is completed, and then performs laser bending forming again, which is time-consuming and laborious. SUMMARY

[0003] The present application proposes a multi-pass flexible flanging forming trajectory optimization method to solve the problem that the existing flexible flanging forming technology is difficult to realize springback compensation after forming because the sheet material is severely work-hardened, resulting in low flanging forming quality. The flanging forming trajectory is continuously updated during the flanging forming process to realize real-time optimization of springback, effectively solving the shortcomings of the existing sheet metal flanging forming technology. The method is simple and feasible, can reduce the number of springback compensation after multi-pass flanging, avoid the problem of difficult compensation of springback after multi-pass flanging, and effectively improve the forming efficiency and forming quality. It has important engineering application value in the fields of aerospace, shipbuilding and automobile manufacturing.

[0004] The present application is realized by the following technical solutions:

[0005] The present application relates to a multi-pass flexible flanging forming trajectory optimization method. The flanging springback compensation criteria are set according to the material of the target flanging part, the multi-pass flexible flanging forming strategy is determined according to the characteristics of the flanging part, and the theoretical flanging forming trajectory and the total forming time are comprehensively obtained. During the actual flanging mechanism forming process, the real-time working condition data are collected by a vision system and imported into a forming trajectory prediction algorithm to predict the forming trajectory at the next moment, realize closed-loop control of the flanging mechanism, and complete the entire flanging forming process.

[0006] The multi-pass flexible flanging forming strategy includes: flanging angle allocation strategy, rotational motion strategy, revolution flanging forming strategy, and rotation-revolution matching strategy.

[0007] The aforementioned flanging angle allocation strategy refers to the flanging angle allocation strategy when the sheet metal is formed in multiple passes according to the characteristics of the part, such as uniform distribution of flanging angle in each pass, gradual increase of flanging angle in each pass, gradual decrease of flanging angle in each pass, and flanging angle allocation strategy according to multiple functions.

[0008] The aforementioned rotational motion strategy refers to the motion strategy of the flanging forming roller when it rotates to the flanging corner of each flanging pass, including but not limited to uniform motion, accelerated motion, and decelerated motion.

[0009] The aforementioned revolution-flanging forming strategy refers to the following: after the forming roller rotates to the target flanging angle of each pass, the entire sheet is flanged. This can be achieved by using auxiliary tools such as robots to drive the forming roller to revolve, or by using a platform on which the sheet is located to rotate while the forming roller remains stationary.

[0010] The aforementioned rotation-revolution matching strategy refers to the matching relationship between the rotation of the forming roller to the target flanging angle in each pass and the revolution of the entire sheet metal for flanging, such as rotating first and then revolving, or rotating and revolving simultaneously at different speeds.

[0011] The aforementioned vision system refers to a laser vision measurement system:

[0012] The real-time operating data includes: real-time edge turning, real-time track number, and real-time roller position.

[0013] The aforementioned trajectory prediction algorithm refers to a recurrent neural network based on processing time series data, including but not limited to RNN, LSTM, and GRU.

[0014] The present application relates to a system for implementing the above method, comprising: a flanging forming decision module, a vision module, a flanging forming real-time optimization module, and a flanging forming execution module, wherein: the flanging forming decision module determines a multi-pass flexible flanging forming strategy according to the characteristics of the target part, obtains a theoretical flanging forming trajectory, and determines a flanging springback compensation criterion according to the material of the part; the vision module collects and records the real-time flanging angle, real-time pass, and real-time roller position during the flanging forming process; the flanging forming real-time optimization module obtains the flanging forming trajectory at the next moment by using the forming trajectory prediction algorithm and the flanging springback compensation criterion determined by the flanging forming decision module, according to the data collected by the vision module and the theoretical flanging forming trajectory specified by the flanging forming decision module; the flanging forming execution module initially performs flanging forming according to the theoretical flanging forming trajectory specified by the flanging decision module, and performs flanging forming according to the flanging forming trajectory at the next moment specified by the flanging forming real-time optimization module, and continuously performs flanging forming according to the flanging forming trajectory at the next moment specified by the flanging forming real-time optimization module, to realize real-time adjustment of flanging forming.

[0015] Technical effects

[0016] The present application realizes real-time springback compensation by dynamically adjusting the trajectory through the flanging forming trajectory prediction algorithm according to the change of the flanging angle during the flanging forming process. Compared with the prior art, the present application is simple and feasible, and changes the existing practice of performing springback compensation after flanging forming is completed. By performing real-time springback compensation during the flanging forming process, the springback compensation pass after flanging is reduced or eliminated, the work hardening degree of the sheet metal is reduced, and the flanging part forming quality and forming efficiency are improved. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 The flowchart of the present application;

[0018] Figure 2 The size diagram of the target flanging part of the embodiment;

[0019] In the figure: 101 is the planar part of the target flanging part, 102 is the transition fillet part of the target flanging part, and 103 is the flanging part of the target flanging part;

[0020] Figure 3 The size diagram of the original sheet metal of the embodiment;

[0021] Figure 4 The forming trajectory comparison diagram of the present application and the existing flanging forming technology;

[0022] Figure 5 The flanging angle comparison diagram of the parts obtained by using the forming trajectory of the present application and the existing flanging forming technology. DETAILED DESCRIPTION

[0023] AsFigure 1 As shown, the present embodiment relates to a multi-pass flexible flanging forming trajectory optimization method, which comprises:

[0024] Step 1) analyze the material performance of the target part, and set the springback compensation criterion: for example Figure 2 and Figure 3 As shown, the material of the target flanging part is LY12 aluminum alloy, so the flanging springback compensation criterion is set to 1 / 2 compensation criterion, that is, half of the reverse compensation springback angle;

[0025] Step 2) determine the multi-pass flexible flanging forming strategy according to the features of the flanging part: according to the thickness of the target flanging part is 2mm, the flanging angle is 60°, the circular opening angle is 90°, the inner diameter of the sheet is 1460mm, the flanging length is 30mm and other features, therefore, set 6-pass flanging forming, the flanging angle distribution strategy is to uniformly distribute the angle of each pass, that is, the flanging angle of each pass is set to 10°, the rotation motion strategy is to select uniform motion, the motion speed is set to 1° / s, the revolution flanging forming strategy is to select the strategy of robot driving the forming roller to rotate to realize the revolution, the motion speed is set to 3° / s, the rotation revolution matching strategy is to select the strategy of robot driving the forming roller to rotate to the target flanging angle of each pass before driving the forming roller to revolve to realize the forming of the entire sheet, and the total forming time can be determined to be 240s; after obtaining the theoretical flanging forming trajectory, start the flanging mechanism for forming while collecting data by using the vision system, set the time when the vision system starts to collect data s=0, and obtain the real-time flanging angle at 0-5 time;

[0026] Step 3) import the theoretical forming trajectory and the real-time flanging angle at 0-5 time into the forming trajectory prediction algorithm based on LSTM, predict the flanging angle at the 6th second based on the historical data at 0-5 time, and obtain the forming trajectory at the 6th second based on the set 1 / 2 springback compensation criterion, and then update the system forming trajectory for flanging forming;

[0027] Step 4) when the forming at the 6th second is completed, import the theoretical forming trajectory and the real-time flanging angle at 1-6 time into the forming trajectory prediction algorithm based on LSTM, predict the flanging angle at the 7th second based on the historical data at 1-6 time, and obtain the forming trajectory at the 7th second based on the set 1 / 2 springback compensation criterion, and then update the system forming trajectory for flanging forming; repeat the above operation until the entire flanging forming process is completed.

[0028] As shown in Figure 4 Compared with the trajectory of the existing flanging forming technology, the flanging forming trajectory of the present application considers the anisotropy of the material or the asymmetry of the part shape, adjusts the forming trajectory according to the different flanging springback, and thereby improves the flanging forming quality.

[0029] As shown in Figure 5As shown in the figure, the part flanging angle obtained by the forming track of the present application and the existing flanging forming technology is compared. Compared with the average flanging springback angle of 1.57° obtained by the existing flanging forming technology, the average springback of the flanging part obtained by the present application is only 0.19°, which significantly improves the forming quality; the flanging part formed by the forming track obtained by the present method has a maximum flanging springback of 0.27°, which is flat and has no obvious defects such as collapse.

[0030] In summary, the present application adjusts the forming track in real time during the flanging forming process, without the need to increase the pass after the flanging forming is completed to compensate for the springback, thereby greatly improving the forming efficiency.

[0031] The above specific embodiments can be adjusted in different ways by those skilled in the art without departing from the principles and purposes of the present application, the protection scope of the present application is subject to the claims and is not limited by the above specific embodiments, and each implementation scheme within the scope is subject to the present application.

Claims

1. A multi-pass flexible flanging forming trajectory optimization method, characterized in that, According to the target flanging material setting flanging springback compensation criterion, combined with the flanging part feature determines the multi-pass flexible flanging forming strategy, comprehensive get the theoretical flanging forming trajectory and total forming time; In the actual flanging mechanism forming process, through the visual system to collect real-time working condition data and import forming trajectory prediction algorithm, predict the next time forming trajectory, realize the closed loop control of flanging mechanism, until the completion of the whole flanging forming process; The multi-pass flexible flanging forming strategy comprises a flanging angle distribution strategy, a rotation motion strategy, a revolution flanging forming strategy and a rotation revolution matching strategy. The flanging angle distribution strategy refers to the distribution strategy of the flanging angle of the sheet metal in the multi-pass forming according to the part feature, The rotation motion strategy refers to the motion strategy of the flanging forming roller when it rotates to the flanging angle of the pass, The revolution flanging forming strategy refers to the flanging forming of the whole sheet metal after the forming roller rotates to the target flanging angle of the pass, The rotation revolution matching strategy refers to the matching relationship between the rotation of the forming roller to the target flanging angle and the revolution flanging forming of the whole sheet metal. The real-time working condition data comprises real-time flanging angle, real-time pass and real-time position of the roller.

2. The multi-pass flexible flange forming trajectory optimization method of claim 1, wherein, The flanging angle distribution strategy comprises uniform distribution of flanging angle of each pass, gradual increase of flanging angle of each pass, gradual decrease of flanging angle of each pass and flanging angle distribution strategy according to multiple functions.

3. The multi-pass flexible flange forming trajectory optimization method of claim 1, wherein, The forming trajectory prediction algorithm refers to a recurrent neural network based on time series data processing.

4. The multi-pass flexible flange forming trajectory optimization method of any one of claims 1-3, wherein, It comprises: Step 1) analyze the material performance of the target part, set the springback compensation criterion: the target flanging part material is LY12 aluminum alloy, so the flanging springback compensation criterion is set to 1 / 2 compensation criterion, that is, half of the reverse compensation springback angle; Step 2) determine the multi-pass flexible flanging forming strategy according to the flanging part feature: according to the target flanging part thickness of 2mm, the flanging angle of 60°, the circular opening angle of 90°, the sheet metal inner diameter of 1460mm, the flanging length of 30mm and other features, set 6-pass flanging forming, the flanging angle distribution strategy is to uniformly distribute the flanging angle of each pass, that is, to set the flanging angle of each pass to 10°, the rotation motion strategy is to select uniform motion, the motion speed is set to 1° / s, the revolution flanging forming strategy is to select the strategy of robot driving the forming roller to realize revolution, the motion speed is set to 3° / s, the rotation revolution matching strategy is to select the strategy of robot driving the forming roller to rotate to the target flanging angle of each pass and then driving the forming roller to revolution to realize the forming of the whole sheet metal, the total forming time can be determined as 240s; after obtaining the theoretical flanging forming trajectory, start the flanging mechanism to form while collecting data by using the visual system, set the visual system to start collecting data at time s=0, and obtain the real-time flanging angle at 0-5 time. Step 3) The theoretical forming trajectory and the real-time flanging angle at 0-5 s are introduced into the LSTM-based forming trajectory prediction algorithm, the flanging angle at 6 s is predicted based on the historical data at 0-5 s, and the forming trajectory at 6 s is obtained based on the set 1 / 2 springback compensation criterion, and then the system forming trajectory is updated for flanging forming; Step 4) After the forming at 6 s is completed, the theoretical forming trajectory and the real-time flanging angle at 1-6 s are introduced into the LSTM-based forming trajectory prediction algorithm, the flanging angle at 7 s is predicted based on the historical data at 1-6 s, and the forming trajectory at 7 s is obtained based on the set 1 / 2 springback compensation criterion, and then the system forming trajectory is updated for flanging forming; the above operation is repeated until the entire flanging forming process is completed.

5. A system for implementing the method of multi-pass flexible flanging trajectory optimization according to any one of claims 1-4, characterized in that, It comprises: a flanging forming decision module, a vision module, a flanging forming real-time optimization module, and a flanging forming execution module, wherein: the flanging forming decision module determines a multi-pass flexible flanging forming strategy according to the characteristics of the target part, obtains a theoretical flanging forming trajectory, and determines a flanging springback compensation criterion according to the material of the part; the vision module collects and records the real-time flanging angle, the real-time pass, and the real-time position of the roller during flanging forming; the flanging forming real-time optimization module obtains the flanging forming trajectory at the next moment according to the data collected by the vision module and the theoretical flanging forming trajectory specified by the flanging forming decision module, using the flanging forming decision module and the flanging springback compensation criterion determined by the flanging forming decision module; The flanging forming execution module initially performs flanging forming according to the theoretical flanging forming trajectory specified by the flanging decision module, and performs flanging forming according to the flanging forming trajectory at the next moment specified by the flanging forming real-time optimization module, and continuously performs flanging forming according to the flanging forming trajectory at the next moment specified by the flanging forming real-time optimization module, realizing real-time adjustment of flanging forming.

Images