CN1353039A - Intelligent Control Method of Injection Molding Machine - Google Patents
Intelligent Control Method of Injection Molding Machine Download PDFInfo
- Publication number
- CN1353039A CN1353039A CN 00133427 CN00133427A CN1353039A CN 1353039 A CN1353039 A CN 1353039A CN 00133427 CN00133427 CN 00133427 CN 00133427 A CN00133427 A CN 00133427A CN 1353039 A CN1353039 A CN 1353039A
- Authority
- CN
- China
- Prior art keywords
- injection molding
- molding machine
- neural network
- quality
- intelligent control
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 46
- 238000001746 injection moulding Methods 0.000 title claims abstract 17
- 238000013528 artificial neural network Methods 0.000 claims abstract description 45
- 238000004519 manufacturing process Methods 0.000 claims abstract description 18
- 238000003050 experimental design method Methods 0.000 claims abstract 2
- 238000012937 correction Methods 0.000 claims description 11
- 238000012360 testing method Methods 0.000 claims description 3
- 238000002347 injection Methods 0.000 claims 1
- 239000007924 injection Substances 0.000 claims 1
- 230000001537 neural effect Effects 0.000 claims 1
- 238000000465 moulding Methods 0.000 abstract description 36
- 239000000463 material Substances 0.000 abstract description 9
- 230000008859 change Effects 0.000 abstract description 2
- 230000008569 process Effects 0.000 description 19
- 230000015572 biosynthetic process Effects 0.000 description 15
- 238000012544 monitoring process Methods 0.000 description 6
- 239000004744 fabric Substances 0.000 description 5
- 238000002474 experimental method Methods 0.000 description 4
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000013401 experimental design Methods 0.000 description 2
- 239000003292 glue Substances 0.000 description 2
- 238000013179 statistical model Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 240000004859 Gamochaeta purpurea Species 0.000 description 1
- 238000010924 continuous production Methods 0.000 description 1
- 230000001955 cumulated effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000005206 flow analysis Methods 0.000 description 1
- 239000012530 fluid Substances 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 239000000155 melt Substances 0.000 description 1
- 238000002844 melting Methods 0.000 description 1
- 230000008018 melting Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000004886 process control Methods 0.000 description 1
- 238000003908 quality control method Methods 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
Images
Landscapes
- Injection Moulding Of Plastics Or The Like (AREA)
Abstract
The intelligent control method of the injection molding machine of the invention applies neural network control and prediction to the injection molding of the injection molding machine, and adopts a product quality multiple loss function as an index for optimizing the intelligent control. The experimental results of the machine are analyzed in the CAE model flow software program according to an experimental design method, and an intelligent control mode and a quality prediction mode are trained according to the experimental results. The invention can quickly set parameters close to the optimal values in a plurality of molding cycles, produce products with the least total loss, increase the production capacity, automatically detect and adjust the change factors of machines, environments and materials, and maintain the production of molded products meeting the quality requirements.
Description
The present invention particularly controls and predicts the method for the ejection formation of Jet forming machine about the method for the intelligent control of Jet forming machine about the application class neural network.
Traditionally, the ejection formed piece quality control can be divided into two stages, and the phase I is set (selected ejaculator input value) for the Jet forming machine parameter.Set the moulding product that can obtain conforming with quality requirements thus.Moulding product quality monitoring and control when second stage is production.There is diverse ways to carry out the ejaculator parameter respectively with regard to this two stage and sets and monitor the moulding quality.
The Jet forming machine parameter is set, as: melting glue temperature, mold temperature, ejaculation pressure and issuing velocity etc. is to penetrate the groundwork that parameter is set.Ideal situation, the input value of Jet forming machine can be produced the moulding product that conform with quality requirements after determining.But because the manufacture process of Jet forming machine is dynamic complicated and frequent the variation, it is very difficult selecting suitable input value.During its employed plastic material melt-flow for have non-newtonian fluid, non-equal thermal characteristics and also in the ejection formation process fused raw material influenced by complicated factors such as high pressure and high temperature.In addition, the dynamic characteristic of Jet forming machine is also very complicated, so the Jet forming machine parameter is set and need further be inquired into.Trial and error pricing commonly used is in the past made the Jet forming machine parameter and is set, and the engineering staff uses them to set the ejaculator parameter in the knowledge on manufacture process, conforms with quality requirements for this reason through the correction of constantly examination mistake up to product.The shortcoming of trial and error pricing maximum is that efficient is not high.The engineering staff only can rely on ejaculator molding parameter and quality when parameter is set be that linear relationship and one time one parameter are set, and therefore can't significantly adjust.What is more, and the engineering staff often sets different machine parameter and the relation of product quality parameter characteristic is considered as independently having ignored association between them.Therefore parameter set very numerous and diverse originally, add machine, environment, and the changing factor of material that machine parameter is set is more difficult.And the parameter set of trial and error pricing is most probably at the edge of the molding parameter bound of product, and product quality can be very unstable in process of production.
Recently this several years expert systems have attracted great notice, and the engineering staff adopts expert system suggestion to make the ejaculator parameter and sets, if its principle be with expertise with one so (if-then) provide the debug rule to solve problem.But expert system has it restricted, as between molding parameter and quality parameter, there is no clear and definite qualitative relationships, and it can't provide the information that exceeds outside its rule.
The more systematic ejaculator parameter method of forming is to use field mouth method or experiment to plan method to obtain enough information and set up an empirical mode, do in the process of ejaculator parameter moulding in pattern according to this, have at least 25 ejaculator molding parameters can influence the quality of moulding product, and also can interact between the moulding ejaculator molding parameter.Therefore need expend considerable time and material resources in experiment and the analytic process, and its result also only can be fit to mutual relation between certain specific board and some specific molding parameter and the quality.
Because between the ejaculator molding parameter of input and the product quality and non-linear relation, and penetrate complicate fabrication process and often change the product quality monitoring and the control that cause ejaculator and become extremely important.In early days, the engineering staff only can be according to experience and range estimation, and irregular inspection product is also adjusted the purpose that the ejaculator parameter reaches the control product according to this.The method has several big shortcomings.At first, the operator will have enough training just can judge moulding product quality accurately.And, some outward appearances and the major defect of the judgement moulding product that visual method only can be rough.So some the non-visual defectives that can judge, as: warpage, size variation and residual stress ... or the like, the problem of just finding when these are everlasting the ejection formed piece assembling often has been left in the basket.
The statistics program mode control method is to measure manufacture process signal and product quality, and sets up a statistical model that penetrates between parameter setting and product quality according to this, for example: nozzle pressure inference product size or weight during with ejection formation.And the shortcoming of statistics manufacture process pattern is need expend in setting up the statistical model process between the molding parameter and quality parameter of considerable time and material resources and moulding product, can't summarize clear and definite qualitative relationships.
The global patent relevant with ejection formation roughly surpasses thousand pieces nearly ten year every year, and rises year by year, shows that ejection forming technique is just flourish.About the global patent of ejection formation manufacture process control, by 1970 to 2000 Christian era record have 173 pieces (data source ep.espacenet.com).The patent more relevant with the present invention is No. the 5997778th, United States Patent (USP), and its content is known issuing velocity curve of input and the dynamic response of trying to achieve ejaculator, with PID feedback control (PID) setup parameter and Sustainable Control ejection formation.No. the 5246645th, United States Patent (USP), United States Patent (USP) No. 5514311 and No. 4820464, its content is the filling pressurize changing method of control ejection formation.No. the 5997778th, United States Patent (USP) No. 48502178 and No. 48116197, uses PVT to concern the quality of controlling the moulding product.No. the 4311446th, United States Patent (USP) uses screw position, speed, melts the glue temperature and known or desired value is relatively reached ejection formation control.No. the 4060362nd, United States Patent (USP), No. the 3893792nd, United States Patent (USP) and patent No. WO9943487 make materials pipe pressure or die cavity pressure feedback control ejection formation parameter.No. the 61118223rd, Japan Patent and Japan Patent are reached ejection formation with the control screw speed No. 63125322 and are controlled.Above-mentioned patent all only can be finished the monitoring and the control of part molding parameter and moulding product quality.And the present invention can solve molding parameter setting and problems such as moulding product quality monitoring and control.
Main purpose of the present invention provides the ejection formation of control of application class neural network and prediction Jet forming machine, in simplation verification, can in several molding cycles, set the ejaculator parameter rapidly and be similar to best setting, and produce the product of minimal loss and increase production capacity, and the Auto-Sensing adjustment overcomes machine, environment, and the changing factor of material, keeps the moulding product that conform with quality of producing.Therefore, can solve at one stroke that the Jet forming machine parameter is set and problems such as moulding product quality monitoring and control when producing, really reach the Based Intelligent Control of Jet forming machine.
For reaching above-mentioned purpose of the present invention, the invention provides a kind of method of intelligent control of Jet forming machine, comprise the following steps: to carry out mould stream software program, and with the comparison that performs an analysis of this Jet forming machine practical test result, in order to set up the data bank of a qualitative relationships, wherein this data bank contains Jet forming machine parameter, process parameters and quality parameter parameter at least; Definition total losses function formula, and, calculate two groups of data of the corresponding quality parameter of the corresponding Jet forming machine parameter of total losses parameter with process parameters according to the data bank of qualitative relationships; According to qualitative relationships and data bank and two groups of data, set up a Based Intelligent Control class neural network and quality prediction class neural network; The output of the Based Intelligent Control class neural network after the Based Intelligent Control class neural network set up, total losses function formula and the quality prediction class neural network coupled in series set up is connected to the input of this Jet forming machine, and the input of the quality prediction class neural network after will coupling is connected to the output of this Jet forming machine; The Based Intelligent Control class neural network that wherein couples receives the total losses that carries out Correction and Control of the total losses function formula output that couples, and output is carried out the Jet forming machine parameter of Correction and Control to this Jet forming machine; The quality prediction class neural network that wherein couples receives the process parameters of carrying out Correction and Control of this Jet forming machine output, and output is carried out the quality parameter parameter of Correction and Control to the total losses function formula that couples.
The ejection formation of application class neural network control of the present invention and prediction Jet forming machine, for making intelligent control energy optimization, this law adopts the multiple loss function of product quality as index.Target of the present invention is to obtain product high-quality, with short production cycle and that production cost is low, promptly produces fast, quality better, low cost.Can divide two big classes with regard to this target, first has minimum total quality loss for the moulding product.Second for increasing output.Promptly target is designed according to this for specific embodiments of the invention, the mode that the present invention reaches this target is after the ejection formed piece design, the engineer uses mould stream software program (CAE), sets up pattern, molding parameter and the selection of material, analyzes output moulding bound and optimal parameter recommended value.Planning method according to experiment again, is level with the higher limit and the lower limit of Jet forming machine parameter, carries out the CAE mold flow analysis and penetrate experiment on the ejaculator board, writes down both Jet forming machine parameter, process parameters and product quality parameter etc. and analyzes the result.Definition total losses function merges the multiple loss of both analysis results and counting yield quality, obtains losing corresponding Jet forming machine parameter and manufactures two groups of data of the corresponding quality parameter of parameter parameter.Again with two groups of data difference training smart control models and quality prediction pattern two class neural network.
The present invention is similar to best setting by the ejaculator parameter of setting rapidly known in the simplation verification in several molding cycles, and produces the product of minimum total quality loss and increase production capacity.And can Auto-Sensing adjustment overcome machine, environment, and the variable of material keep the moulding product that conform with quality requirements of producing.Therefore, the present invention can finish the setting of Jet forming machine parameter and the moulding product that automatic continuous production conforms with quality requirements at one stroke; The present invention and applicable on different moulds and the type.
The above-mentioned method of experimental design of the present invention can utilize existing field mouth formula method of experimental design (Taguchi Parameter Design Method) mould stream software program (CAE) can utilize existing U.S. Cornell (Comell) the C-MOLD mold flow analysis software program that university developed.
Fig. 1 is the specific embodiment of the present invention on a Jet forming machine;
Fig. 2 penetrates the basic function neural network for the width of cloth that the present invention uses.
Fig. 3 is the method flow of the intelligent control of the present invention.
The specific embodiment of Jet forming machine control method of the present invention as shown in Figure 1.So that Jet forming machine begins ejection formation among Fig. 1, after one group of intelligent control class neural network input earlier is positioned at the setting value of parameter bound, signal in the manufacture process of ejection formation is measured and arrangement by sensor, predict the quality of this molding cycle product via quality prediction class neural network, calculate the total losses of this molding cycle product through the total losses function formula.Judge by intelligent control class neural network again and export next molding cycle Jet forming machine parameter and give cyclelog that Jet forming machine carries out ejection formation of following one-period according to this parameter.This control flow goes round and begins again, and can reach the targets such as Automatic parameter setting, quality monitoring and control of Jet forming machine.
The intelligent control class neural network, the total losses that further specify are as shown in Figure 1 contained number formula and quality prediction neural network.Data bank and two groups of data according to above-mentioned qualitative relationships, set up a Based Intelligent Control class neural network and quality prediction networking, and the Based Intelligent Control class neural network that will set up, total losses function formula and the quality prediction class neural network coupled in series of having set up, and the output of the Based Intelligent Control class neural network after will coupling is connected to the input of this Jet forming machine, and the input of the quality prediction class neural network after will coupling is connected to the output of this Jet forming machine, the Based Intelligent Control class neural network that wherein couples receives the total losses that carries out Correction and Control of the total losses function formula output that couples, and output is carried out the Jet forming machine parameter of Correction and Control to this Jet forming machine, receive the process parameters of carrying out Correction and Control of this Jet forming machine output with the quality prediction class neural network that wherein couples, and export and carry out the quality parameter parameter of Correction and Control to the total losses function formula that couples.In concrete enforcement, above-mentioned Jet forming machine parameter comprises issuing velocity, dwell time, dwell pressure at least.Above-mentioned process parameters comprises filling velocity, dwell time, die cavity pressure at least.Above-mentioned quality parameter parameter comprises that at least output weight, maximum volume shrink, maximum depression.
Fig. 2 penetrates the basic function neural network for the width of cloth that the present invention uses.Based Intelligent Control class neural network that the present invention is above-mentioned and quality prediction class neural network adopt two width of cloth to penetrate the basic function neural network at specific embodiment.A width of cloth is penetrated the basic function neural network and is implemented as Based Intelligent Control class neural network, and basic output layer is the output of Based Intelligent Control class neural network, output Jet forming machine parameter.Another width of cloth is penetrated the basic function neural network and is implemented as quality prediction class neural network, its output layer is the output of quality prediction class neural network, output quality parameter parameter is to the total losses function formula, another input layer is the input of quality prediction class neural network, and it receives the process parameters of Jet forming machine output.
Fig. 3 cumulated volume is invented above disclosed content, represents the flow process of the method for intelligent control of the present invention.
Though the present invention with a preferred embodiment openly as above; yet it is not in order to limit the present invention; anyly be familiar with this operator; without departing from the spirit and scope of the present invention; when can doing various changes and modification, so protection scope of the present invention should be as the criterion with the protection domain that accompanying Claim was defined.
Claims (6)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 00133427 CN1353039A (en) | 2000-11-03 | 2000-11-03 | Intelligent Control Method of Injection Molding Machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 00133427 CN1353039A (en) | 2000-11-03 | 2000-11-03 | Intelligent Control Method of Injection Molding Machine |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1353039A true CN1353039A (en) | 2002-06-12 |
Family
ID=4595710
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 00133427 Pending CN1353039A (en) | 2000-11-03 | 2000-11-03 | Intelligent Control Method of Injection Molding Machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1353039A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101637964A (en) * | 2008-07-30 | 2010-02-03 | 鸿富锦精密工业（深圳）有限公司 | Projecting machine system and setting method of parameter thereof |
| CN101430720B (en) * | 2007-11-09 | 2011-02-09 | 邱维铭 | Analytical methods for in-mold decoration injection molding |
| CN104552846A (en) * | 2013-10-25 | 2015-04-29 | 延锋伟世通金桥汽车饰件系统有限公司 | Method for setting mould according to deformation quantity of glass fibre product |
| CN106154979A (en) * | 2015-04-09 | 2016-11-23 | 弘讯科技股份有限公司 | Program design apparatus and method of writing program |
| CN111231251A (en) * | 2020-01-09 | 2020-06-05 | 杭州电子科技大学 | Product detection method, equipment and system of injection molding machine |
-
2000
- 2000-11-03 CN CN 00133427 patent/CN1353039A/en active Pending
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101430720B (en) * | 2007-11-09 | 2011-02-09 | 邱维铭 | Analytical methods for in-mold decoration injection molding |
| CN101637964A (en) * | 2008-07-30 | 2010-02-03 | 鸿富锦精密工业（深圳）有限公司 | Projecting machine system and setting method of parameter thereof |
| CN104552846A (en) * | 2013-10-25 | 2015-04-29 | 延锋伟世通金桥汽车饰件系统有限公司 | Method for setting mould according to deformation quantity of glass fibre product |
| CN106154979A (en) * | 2015-04-09 | 2016-11-23 | 弘讯科技股份有限公司 | Program design apparatus and method of writing program |
| CN106154979B (en) * | 2015-04-09 | 2019-06-28 | 弘讯科技股份有限公司 | Program design apparatus and method of writing program |
| CN111231251A (en) * | 2020-01-09 | 2020-06-05 | 杭州电子科技大学 | Product detection method, equipment and system of injection molding machine |
| CN111231251B (en) * | 2020-01-09 | 2022-02-01 | 杭州电子科技大学 | Product detection method, equipment and system of injection molding machine |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP1166994A1 (en) | Injection moulding control system | |
| US20020019674A1 (en) | Intelligent control method for injection machine | |
| EP1439046B1 (en) | Automated molding technology for thermoplastic injection molding | |
| US5914884A (en) | Method for evaluating moldability characteristics of a plastic resin in an injection molding process | |
| US8983878B2 (en) | Computer-implemented method for optimizing an injection-molding process for producing thick-walled components | |
| CN105636757B (en) | Method for evaluating process-technical properties of injection molds | |
| JPH05507659A (en) | Injection molding machine control method and device | |
| CN113733505B (en) | Injection molding system, molding condition correction system, and injection molding method | |
| CN112770890B (en) | Method of controlling a machine for processing plastics | |
| CN118780697A (en) | An intelligent recommendation method and system for injection mold production process parameters | |
| CN113752505A (en) | Auxiliary device for determining molding conditions and resin state estimation device | |
| CN100405376C (en) | Method for Determining Process Parameters of Plastic Injection and Injection Molding Machine | |
| CN100535908C (en) | Injection molding simulation apparatus and method of injection molding simulation | |
| CN109614651A (en) | A kind of high-precision evaluation method of moulding machined parameters and deformation relationship | |
| CN112659501B (en) | Method for validating a theoretical value curve | |
| CN1353039A (en) | Intelligent Control Method of Injection Molding Machine | |
| Rehmer et al. | A digital twin for part quality prediction and control in plastic injection molding | |
| Chen et al. | Injection molding quality control by integrating weight feedback into a cascade closed‐loop control system | |
| Chang et al. | Cross-machine predictions of the quality of injection-molded parts by combining machine learning, quality indices, and a transfer model | |
| CN114889076A (en) | Injection molding visualization method and device, computer equipment and storage medium | |
| Wang et al. | A concurrent-engineering approach toward the online adaptive control of injection molding process | |
| CN112677438B (en) | Segmented construction and screening method and system based on injection molding and machine debugging data characteristics | |
| Petrova et al. | Incorporation of phenomenological models in a hybrid neural network for quality control of injection molding | |
| Mehta et al. | Performance analysis of plastic injection molding using particle swarm based modified sequential quadratic programming algorithm multi objective optimization model | |
| CN118742432A (en) | Computer-implemented method and system for determining at least one machine |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C06 | Publication | ||
| PB01 | Publication | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |