LU501666B1 - An Intelligent Deep Drawing Control System for Box Parts of Tailor Rolled Blanks - Google Patents

An Intelligent Deep Drawing Control System for Box Parts of Tailor Rolled Blanks Download PDF

Info

Publication number
LU501666B1
LU501666B1 LU501666A LU501666A LU501666B1 LU 501666 B1 LU501666 B1 LU 501666B1 LU 501666 A LU501666 A LU 501666A LU 501666 A LU501666 A LU 501666A LU 501666 B1 LU501666 B1 LU 501666B1
Authority
LU
Luxembourg
Prior art keywords
deep drawing
blank holder
control
tailor rolled
hydraulic
Prior art date
Application number
LU501666A
Other languages
German (de)
Inventor
Huawei Zhang
Xingang Wang
Jialu Wu
Qiuming Li
Original Assignee
Univ Guangdong Petrochem Tech
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Univ Guangdong Petrochem Tech filed Critical Univ Guangdong Petrochem Tech
Priority to LU501666A priority Critical patent/LU501666B1/en
Application granted granted Critical
Publication of LU501666B1 publication Critical patent/LU501666B1/en

Links

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B21—MECHANICAL METAL-WORKING WITHOUT ESSENTIALLY REMOVING MATERIAL; PUNCHING METAL
    • B21D—WORKING OR PROCESSING OF SHEET METAL OR METAL TUBES, RODS OR PROFILES WITHOUT ESSENTIALLY REMOVING MATERIAL; PUNCHING METAL
    • B21D35/00—Combined processes according to or processes combined with methods covered by groups B21D1/00 - B21D31/00
    • B21D35/002—Processes combined with methods covered by groups B21D1/00 - B21D31/00
    • B21D35/005—Processes combined with methods covered by groups B21D1/00 - B21D31/00 characterized by the material of the blank or the workpiece
    • B21D35/006—Blanks having varying thickness, e.g. tailored blanks
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B11/00—Automatic controllers
    • G05B11/01—Automatic controllers electric
    • G05B11/36—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential
    • G05B11/42—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential for obtaining a characteristic which is both proportional and time-dependent, e.g. P. I., P. I. D.
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
    • G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
    • G05B13/0265—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion
    • G05B13/027—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion using neural networks only
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B21—MECHANICAL METAL-WORKING WITHOUT ESSENTIALLY REMOVING MATERIAL; PUNCHING METAL
    • B21D—WORKING OR PROCESSING OF SHEET METAL OR METAL TUBES, RODS OR PROFILES WITHOUT ESSENTIALLY REMOVING MATERIAL; PUNCHING METAL
    • B21D22/00—Shaping without cutting, by stamping, spinning, or deep-drawing
    • B21D22/20—Deep-drawing
    • B21D22/22—Deep-drawing with devices for holding the edge of the blanks

Landscapes

  • Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Evolutionary Computation (AREA)
  • Mechanical Engineering (AREA)
  • Health & Medical Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Feedback Control In General (AREA)

Abstract

The present invention is related to the technical field of tailor rolled blank, and provides an intelligent deep drawing control system for box parts of tailor rolled blanks, which comprises a process monitoring module, an identification and prediction module, a real-time control module and a system execution module. It can complete the real-time monitoring, identification, prediction and control of the deep drawing process of box parts of tailor rolled blanks, to realize the intelligent control of the deep drawing of box parts of tailor rolled blanks, and produce high-quality box parts of tailor rolled blanks. The technical scheme of the present invention can lay a process and technical foundation for the practical manufacturing of large and complex deep drawing parts of tailor rolled blanks, and provide a new idea for improving the forming performance and production efficiency of parts of tailor rolled blanks.

Description

BL-5448 ' LU501666
AN INTELLIGENT DEEP DRAWING CONTROL SYSTEM FOR BOX PARTS OF
TAILOR ROLLED BLANKS Technical Field The present invention relates to the technical field of tailor rolled blanks, in particular to a control system for intelligent deep drawing of box parts of tailor rolled blanks. Background Art Tailor rolled blank is a kind of structural lightweight plate produced by flexible rolling technology. Tailor rolled blanks are different in sheet thickness and mechanical properties. Cracking, wrinkling and transition zone movement during deep drawing will cause uneven wear and shorter service life of the die. This has seriously hindered the application of tailor rolled blanks in the field of vehicle weight reduction. Box parts are representative among non-axisymmetric parts, but are difficult to be shaped. When applied to box parts, the tailor rolled blanks may also face the movement of thickness transition zone caused by the uneven sheet thickness and mechanical properties, in addition to the other most prominent problems in deep drawing such as cracking and wrinkling. Due to the different thickness and properties of the thin and thick side sheets of tailor rolled blanks, coupled with the movement of the transition zone, the deformation process of tailor rolled blanks is more complex, and the uneven deformation of the thin and thick sides will make the shaping of the whole part more difficult. Summary of the Invention The present invention discloses an intelligent deep drawing control system for box parts of tailor rolled blanks, which is characterized in that it comprises a process monitoring module, an identification and prediction module, a real-time control module and a system execution module connected in turn; The process monitoring module monitors the deep drawing process variables in real time with the data acquisition system, including deep drawing force, blank holder force and deep drawing stroke;
BL-5448 LU501666 The identification and prediction module builds a neural network model to obtain the material performance parameters, the optimal blank holder force and the optimal deep drawing speed of tailor rolled blanks; The real-time control module, according to the optimal figures, controls the blank holder force and the deep drawing speed at different positions with the neural network PID control system based on grey prediction; The system execution module executes the deep drawing of tailor rolled blanks.
The present invention discloses a more advanced technical scheme, which realizes progress in the real-time monitoring, identification, prediction and control of the deep drawing of box parts of tailor rolled blanks, and can produce deep drawing parts of tailor rolled blanks with higher quality.
Description of the Drawings To better describe the embodiment of the present invention or the technical scheme of the prior art, a brief introduction of the accompanying drawings to be used in the descriptions of the embodiment or the prior art is made hereby.
Obviously, the drawings below are only the embodiment of the present invention, and for those ordinarily skilled in the art, other drawings based on such drawings can be obtained without making creative endeavors.
Fig. 1 is a schematic diagram of the system module connection of the present invention.
Fig. 2 is a structure diagram of the neural network PID control system based on grey prediction.
Fig. 3 is a working principle diagram of the system of the present invention.
Embodiment The technical scheme in the embodiment of the present invention will be clearly and completely described in combination with drawings of the embodiment of the present invention.
Based on the embodiment of the present invention, all the other embodiments obtained by those ordinarily skilled in the art without making creative endeavors shall fall into the scope of protection of the present invention.
The embodiment discloses an intelligent deep drawing control system for box parts of tailor rolled blanks, which comprises a process monitoring module, an identification and prediction module, a real-time control module and a system execution module, as shown in Fig. 1.
BL-5448 LU501666 The data acquisition system comprises an A/D conversion card, an industrial personal computer, a voltage limiter, a dynamic strain meter, sensors and an acquisition, recording and display program based on Lab VIEW. The dynamic strain meter is connected with multi-sensors at the same time for the measurement of dynamic stress and strain, and can realize the functions of signal method and filtering. The sensors include linear displacement sensor and strain gauge pressure sensor whose signals are amplified by the dynamic strain meter. By calibrating the pressure sensor and displacement sensor, the corresponding relationship between sensor voltage and deep drawing force, blank holder force and deep drawing stroke is determined. The sensor converts the deep drawing process variables into analog voltage signals according to the relationship. The A/D conversion card receives and converts the analog voltage signals into digital signals, which will be entered into and stored in the industrial personal computer as data files. The industrial personal computer calculates the deep drawing speed - stroke curve based on the differential relation between speed and displacement, which will be presented together with deep drawing force - stroke curve and blank holder force - stroke curve. The voltage limiter is placed between the dynamic strain meter and the A/D conversion card to limit the voltage within a certain range and protect the system.
The identification and prediction module is mainly composed of computer and program. It identifies the material performance parameters and predicts the optimal technological parameters with the Matlab neural network model optimized on genetic algorithm. The optimal technological parameters are the optimal blank holder force and the optimal deep drawing speed. The LabVIEW main program in the control system analyzes and processes the characteristic information of the machined object obtained by the process detection module, identifies the material performance parameters of the machined object in real time through the neural network identification model, and offers the optimal variable blank holder force curve and the optimal deep drawing speed curve with the neural network prediction model.
The neural network identification model, with the deep drawing process variables and the geometric parameters of the tailor rolled blanks as the input layer parameters and the material performance parameters of the tailor rolled blanks as the output layer parameters, establishes the nonlinear mapping relationship between the material performance parameters of the tailor rolled blanks and the deep drawing process variables and geometric parameters of the tailor rolled blanks; the neural network prediction model, with the material performance parameters, geometric parameters and deep drawing stroke of the tailor rolled blanks as the input layer variables, and the optimal blank holder force and optimal deep drawing speed as the output layer variables,
BL-5448 LU501666 establishes the nonlinear mapping relationship between the optimal blank holder force and optimal deep drawing speed and the material performance parameters, geometric parameters and deep drawing stroke of the tailor rolled blanks. The real-time control module mainly comprises a D/A conversion card, a control program, a proportional overflow valve, a proportional amplifier and an electro-hydraulic proportional valve. The real-time control module is developed on LabVIEW software platform, and the neural network PID control system is established with the neural network PID algorithm based on grey prediction. The said neural network PID control system, with the deep drawing stroke as the feedback signal and the optimal blank holder force and the optimal deep drawing speed as the control signal, obtains the relation curve between the blank holder force value and the control voltage of the proportional overflow valve by calibrating the proportional overflow valve, obtains the relation curve between the deep drawing speed value and the control voltage of the electro-hydraulic proportional valve by calibrating the electro-hydraulic proportional valve, and converts the blank holder force value and the deep drawing speed value into the control voltage signals according to the curves. The digital quantity of the control voltage signals is converted into analog quantity by the D/A conversion card, which is then converted into control current signals by the proportional amplifier to control the opening degree of the proportional overflow valve and electro-hydraulic proportional valve. The pressure of the hydraulic cylinder of the blank holder and the speed of the main cylinder of the hydraulic press will change accordingly, and so will the blank holder force and deep drawing speed.
The system execution module comprises hydraulic control parts and deep drawing parts. In the deep drawing part, multiple sets of blank holder cylinders can be arranged according to the blank holder force required at different positions of the tailor rolled blanks. Each set of blank holder cylinders corresponds to a set of hydraulic control part and deep drawing part. Fig. 3 shows the schematic diagram of four sets of blank holder cylinders; the first pressure sensor and the first displacement sensor are used to monitor the deep drawing force and obtain the deep drawing stroke, the second displacement sensor is used to measure the overall displacement of the blank holder unit, and the second pressure sensor measures the real-time blank holder force provided by the blank holder cylinder.
To realize the variable blank holder force control by areas, blank holder force control units corresponding to the control areas need to be set on the deep drawing die in the hydraulic control parts. The electromagnetic directional valve in the hydraulic control part is to control the up and down movement of the piston of the blank holder cylinder, the throttle valve is to control the
BL-5448 ) LU501666 movement speed of the piston of the blank holder cylinder, and the hydraulic control check valve is to ensure the forward flow of oil. The neural network PID control system completes the process action and blank holder force adjustment through the hydraulic control parts, and completes the deep drawing by controlling the hydraulic control parts and coordinating with the deep drawing parts. The movement speed of the hydraulic press main slider is changed by adjusting the electro- hydraulic proportional valve in the real-time control module.
According to the requirements of input variables of neural network model, the geometric parameters and deep drawing process variables of tailor rolled blanks are input into the network model in real time. The Matlab script program embedded in the framework is the real-time identification neural network model with sample trained, and the output variables of the script framework are the material parameters. In the same way, the material performance parameters, geometric parameters and deep drawing stroke of the tailor rolled blanks are input into the trained neural network model in the script in real time, to predict the optimal blank holder force and the optimal deep drawing speed, and establish the optimal prediction curve.
Referring to Fig. 2, the neural network PID control system is composed of three parts. The first part is an independent incremental PID closed-loop control network composed of optimal blank holder force and optimal deep drawing speed input, PID controller network, proportional valve and sensor. The second part is BP neural network, with inputs of the optimal parameter setting value, grey prediction value and their errors, and outputs of the three parameters Kp, Ki and Ko of PID controller network. For the BP neural network, the error function of each control loop and the error function between loops are selected as the performance index functions, the non- negative S function is used as the activation function of the output node, the positive and negative symmetrical S function is used as the activation function of hidden layer neurons, and the gradient descent method is used for the correction of network weighting coefficient. The third part is the grey prediction model, which takes the optimal blank holder force value, the optimal deep drawing speed value, and the output values of the proportional overflow valve and electro-hydraulic proportional valve as the inputs, and the grey prediction value as the output. The grey prediction model realizes advance control by the following principle: the output blank holder force p(t) of the current time t is obtained from the sensor, which forms a group of data sequence with the data before time t; there is another group of input data sequence r(t); a GM(1,N) model can be established with the two groups of data; the output data y(t+1) of the next time is predicted by the model; y(t+1) replaces p(t) in the actual output and is compared with the input r(t+1) of time t+1 to get the error Ae(t+1).
BL-5448 LU501666 During the no-load pressing of the die, the voltage of the proportional overflow valve 1s changed for many times, to measure the blank holder force values at different voltage values, and obtain the relation curve between the control voltage of the proportional overflow valve and the blank holder force. The functional relationship between them is determined after fitting, and written into the experimental program, so as to drive the voltage regulation of the proportional overflow valve by the blank holder force value, change the opening degree of the proportional overflow valve to control the pressure of the hydraulic cylinder, and thereby control the blank holder force. In the same way, the relationship between the deep drawing speed and the control voltage of the electro-hydraulic proportional valve is determined, to ensure that the deep drawing speed changes according to the optimal prediction curve. The real-time control device can realize the blank holder force control linked with the stroke and position. The blank holder force varying with the stroke can be set through the blank holder force - stroke curve, and the blank holder force varying with the position can be realized by setting different blank holder force - stroke curves for different blank holder cylinders.
The blank holder cylinder drives the blank holder upward. The position of the blank holder is fed back to the control software of the industrial personal computer by the second displacement sensor. After reaching the preset blank holder force zero position, the industrial personal computer sends a control command to power off the upward valve, to make the blank holder in-place at the preset position. After the blank holder is in place, the tailor rolled blank is placed between the blank holder and the concave die. The control software of the industrial personal computer sends a control command to make the hydraulic press main slider go down, which initiates the deep drawing action. In the deep drawing stage, the first displacement sensor feeds back the deep drawing stroke to the industrial personal computer, and the second pressure sensor and the first pressure sensor transmit the blank holder force and the deep drawing force respectively to the industrial personal computer in real time. The real-time feedback control of the deep drawing speed and blank holder force in each area is realized through the control algorithm, ensuring the synchronous output of deep drawing speed and blank holder force according to the optimal prediction curve until the end of drawing.
The above description of the disclosed embodiment enables those skilled in the art to practice or use the present invention. Various modifications to the embodiment will be apparent to those skilled in the art. Accordingly, the present invention will not be limited to the embodiment described herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims (4)

BL-5448 LU501666 CLAIMS
1. An intelligent deep drawing control system for box parts of tailor rolled blanks, which 1s characterized in that it comprises a process monitoring module, an identification and prediction module, a real-time control module and a system execution module connected in turn; The process monitoring module monitors the deep drawing process variables in real time with the data acquisition system, including deep drawing force, blank holder force and deep drawing stroke; The identification and prediction module builds a neural network model to obtain the material performance parameters, the optimal blank holder force and the optimal deep drawing speed of tailor rolled blanks; The real-time control module, according to the optimal figures, controls the blank holder force and the deep drawing speed at different positions with the neural network PID control system based on grey prediction; The system execution module executes the deep drawing of tailor rolled blanks.
2. An intelligent deep drawing control system for box parts of tailor rolled blanks according to Claim 1, which is characterized in that the data acquisition system comprises an A/D conversion card, an industrial personal computer, a voltage limiter, a dynamic strain meter and a sensor; The dynamic strain meter is connected to the sensor. The sensor converts the deep drawing process variables into analog voltage signals. The A/D conversion card receives and converts the analog voltage signals into digital signals, which will be entered into the industrial personal computer for the deep drawing speed - stroke curve. The voltage limiter is placed between the dynamic strain meter and the A/D conversion card.
3. An intelligent deep drawing control system for box parts of tailor rolled blanks according to Claim 1, which is characterized in that the real-time control module comprises a D/A conversion card, a proportional overflow valve, a proportional amplifier and an electro-hydraulic proportional valve. The said neural network PID control system, with the deep drawing stroke as the feedback signal and the optimal blank holder force and the optimal deep drawing speed as the control signal, obtains the relation curve between the blank holder force value and the control voltage of the proportional overflow valve by calibrating the proportional overflow valve, obtains the relation curve between the deep drawing speed value and the control voltage of the electro-hydraulic
BL-5448 LU501666 proportional valve by calibrating the electro-hydraulic proportional valve, and converts the blank holder force value and the deep drawing speed value into the control voltage signals according to the relation curves. The digital quantity of the control voltage signals is converted into analog quantity by the D/A conversion card, which is then converted into control current signals by the proportional amplifier.
4. An intelligent deep drawing control system for box parts of tailor rolled blanks according to Claim 1, which is characterized in that the system execution module comprises hydraulic control parts and deep drawing parts; The hydraulic control parts correspond to different blank holder force control units according to the optimal blank holder force and the optimal deep drawing speed required by each area. Each of the different blank holder force control units comprises a motor, a plunger pump, an oil tank, a hydraulic cylinder, a filter, an electromagnetic directional valve, a throttle valve and a hydraulic control check valve. The motor pumps the oil out from the oil tank through the plunger pump, which passes through the filter, the hydraulic control check valve and the throttle valve into the electromagnetic directional valve, realizing the forward and backward flow of oil. The oil finally enters the hydraulic cylinder; The deep drawing part comprises a deep drawing die, a hydraulic press main slider, a concave die, a convex die, a first pressure sensor, a second pressure sensor, a first displacement sensor, a second displacement sensor, a blank holder cylinder, a blank holder, a lower template piston rod and a blank holder block; the deep drawing die is inverted, the first pressure sensor and the first displacement sensor are installed between the hydraulic press main slider and the concave die, the second displacement sensor is installed at the bottom of the blank holder unit, the blank holder cylinder and the convex die are placed on the lower template together, and the second pressure sensor is installed between the head of the piston rod of the lower template and the blank holder block.
LU501666A 2022-03-15 2022-03-15 An Intelligent Deep Drawing Control System for Box Parts of Tailor Rolled Blanks LU501666B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
LU501666A LU501666B1 (en) 2022-03-15 2022-03-15 An Intelligent Deep Drawing Control System for Box Parts of Tailor Rolled Blanks

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
LU501666A LU501666B1 (en) 2022-03-15 2022-03-15 An Intelligent Deep Drawing Control System for Box Parts of Tailor Rolled Blanks

Publications (1)

Publication Number Publication Date
LU501666B1 true LU501666B1 (en) 2022-09-15

Family

ID=83231039

Family Applications (1)

Application Number Title Priority Date Filing Date
LU501666A LU501666B1 (en) 2022-03-15 2022-03-15 An Intelligent Deep Drawing Control System for Box Parts of Tailor Rolled Blanks

Country Status (1)

Country Link
LU (1) LU501666B1 (en)

Similar Documents

Publication Publication Date Title
CN1181332C (en) Edge pressing force and punching speed changeable single moving sheet hydraulic test machine
CN104191643B (en) The monitoring of single action hydraulic press drawing velocity and pressure-pad-force and control system and method
CN110788165B (en) Rapid calibration control system and control method for roll gap of seven-roll straightener
CN109158430A (en) A kind of corrugation rolls rolling mill hydraulic servo-system position compensation control method
CN109332686B (en) Floating forming electrohydraulic control system and control method of titanium electrode hydraulic press
US20220008979A1 (en) Variable pulsating, gap control, auto-learning press cushion device
CN112692147B (en) Intelligent drawing control system and method for rolled differential-thickness plate box-shaped piece
JPS61232100A (en) Method and apparatus for setting press condition for stable forming
CN110259769A (en) The electrohydraulic control system and forming technology of 3000T liquid-state die forging hydraulic press
LU501666B1 (en) An Intelligent Deep Drawing Control System for Box Parts of Tailor Rolled Blanks
CN106958548B (en) A kind of cold bed traversing control method synchronous based on ratio
CN101704060B (en) Self-adaptive pre-control method for movable beam position of oversize water press
CN104551838A (en) Combined spring deformation compensating device
CN113941624B (en) Bending forming method, device and system for plate material and computer readable storage medium
CN102873106B (en) Rapid and accurate control method for elongation rate of temper mill
CN112776413B (en) Position pressure hybrid optimization control method of servo press
CN108515133A (en) A kind of double acting hydraulic forging press of multistation
CN2632368Y (en) Precision numerical control stretching hydropress shaping correcting controller
CN209439440U (en) The floating of Ti electrode hydraulic press forms electrohydraulic control system
CN112692169B (en) Deep drawing forming die for box-shaped part of rolled differential thick plate
CN113399475B (en) Rolling pressure measuring method for rolling mill
CN110757883B (en) Accurate motion control method based on position pressure automatic compensation
CN112976668B (en) Feedforward compensation control system and method for improving passive leveling system of hydraulic machine
CN115234528B (en) High-precision double closed-loop control system and control method for stretcher
CN113757221A (en) Liquid supply speed and position open-loop control method

Legal Events

Date Code Title Description
FG Patent granted

Effective date: 20220915