CN115409270A - Neural network model based product production prediction method - Google Patents
Neural network model based product production prediction method Download PDFInfo
- Publication number
- CN115409270A CN115409270A CN202211072340.6A CN202211072340A CN115409270A CN 115409270 A CN115409270 A CN 115409270A CN 202211072340 A CN202211072340 A CN 202211072340A CN 115409270 A CN115409270 A CN 115409270A
- Authority
- CN
- China
- Prior art keywords
- neural network
- data
- production
- prediction
- model
- 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
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 92
- 238000000034 method Methods 0.000 title claims abstract description 63
- 238000003062 neural network model Methods 0.000 title claims abstract description 20
- 238000013528 artificial neural network Methods 0.000 claims abstract description 71
- 230000008569 process Effects 0.000 claims abstract description 33
- 238000012545 processing Methods 0.000 claims abstract description 26
- 238000001746 injection moulding Methods 0.000 claims abstract description 25
- 238000012544 monitoring process Methods 0.000 claims abstract description 14
- 230000009977 dual effect Effects 0.000 claims abstract description 6
- 230000002068 genetic effect Effects 0.000 claims description 31
- 238000004422 calculation algorithm Methods 0.000 claims description 18
- 210000000349 chromosome Anatomy 0.000 claims description 17
- 230000005540 biological transmission Effects 0.000 claims description 14
- 230000007547 defect Effects 0.000 claims description 10
- 238000004088 simulation Methods 0.000 claims description 10
- 230000002950 deficient Effects 0.000 claims description 9
- 238000001816 cooling Methods 0.000 claims description 8
- 238000004458 analytical method Methods 0.000 claims description 7
- 230000006870 function Effects 0.000 claims description 6
- 230000035772 mutation Effects 0.000 claims description 6
- 238000010187 selection method Methods 0.000 claims description 6
- 238000011156 evaluation Methods 0.000 claims description 5
- 238000005206 flow analysis Methods 0.000 claims description 5
- 239000000463 material Substances 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 230000008859 change Effects 0.000 claims description 4
- 238000000465 moulding Methods 0.000 claims description 4
- 238000012549 training Methods 0.000 claims description 3
- 108090000623 proteins and genes Proteins 0.000 claims description 2
- 238000003466 welding Methods 0.000 claims description 2
- 238000007405 data analysis Methods 0.000 abstract description 4
- 238000004891 communication Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 239000000243 solution Substances 0.000 description 3
- 239000002699 waste material Substances 0.000 description 3
- 238000012356 Product development Methods 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000004904 shortening Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/04—Manufacturing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Biology (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Physiology (AREA)
- Genetics & Genomics (AREA)
- Manufacturing & Machinery (AREA)
- Primary Health Care (AREA)
- Injection Moulding Of Plastics Or The Like (AREA)
Abstract
The invention discloses a prediction method for product production based on a neural network model, which relates to the technical field of product production prediction.A data monitoring module is used for acquiring all production data in the injection molding production process of a thin-shell plastic part in real time, a dual neural network prediction unit is established in a data processing module, and all the production data are used as initial parameters of the dual neural network prediction unit to obtain two neural network prediction models; and inputting all production data serving as initial parameters into two neural network prediction models to obtain two prediction results, setting a final judgment module, combining the first prediction result and the second prediction result by a processor to obtain an optimal process parameter combination, and guiding the production process of the product by using the optimal process parameter combination. The invention relates to a prediction method for product production based on a neural network model, which has the advantages of more accurate data analysis result, manpower resource saving and economic cost saving.
Description
Technical Field
The invention relates to the technical field of product production prediction, in particular to a prediction method for product production based on a neural network model.
Background
In the field of automobile manufacturing, a certain amount of matched thin-shell plastic parts need to be produced, the thin-shell plastic parts are usually produced in an injection molding mode in the production process, the injection molding process in the prior art is a multivariable and nonlinear melt flow change process, CAE modeling of a pouring system and a cooling system is carried out on the thin-shell plastic parts according to the shape and material attributes of the thin-shell plastic parts and data recommended by a Moldflow database, and molding process parameters are preliminarily set according to the recommendation, so that a whole set of injection molding numerical simulation system is established, preliminary defect analysis such as warping deformation and weld marks is carried out on the thin-shell plastic parts, manual judgment is carried out according to analysis results, and the production of subsequent products is guided.
The existing data processing system for product production can only collect various data in the production flow, and then guides the subsequent product production through manual analysis and processing, so that the defects of inaccurate data analysis result, waste of human resources and easy error in the process of guiding the product production exist.
Disclosure of Invention
The invention mainly aims to provide a prediction method for product production based on a neural network model, which can effectively overcome the defects of inaccurate data analysis result, waste of human resources and easiness in making mistakes in the process of guiding the product production in the conventional data processing system for product production in the background technology.
In order to achieve the purpose, the invention adopts the technical scheme that: a prediction method for product production based on a neural network model comprises the following steps:
s1: setting a data monitoring module, wherein the data monitoring module acquires all production data in the injection molding production process of the thin-shell plastic part in real time;
s2: the production monitoring system comprises a data transmission module and a data processing module, wherein the data transmission module is mainly used for transmitting all production data to the data processing module, the data processing module is mainly used for processing all production data, the data monitoring module is connected with the data processing module through the data transmission module, the data transmission module is specifically a communication network, the data transmission is realized through the communication network, and meanwhile, the stability of data transmission is ensured;
s3: establishing a dual neural network prediction unit in the data processing module, and establishing a BP neural network prediction model and a GA-BP neural network prediction model in the dual neural network prediction unit;
s4: all production data are used as initial parameters to be input into a BP neural network prediction model, learning, parameter adjustment, threshold value matching and result judgment of the BP neural network prediction model are completed to obtain an optimal BP neural network prediction model, and the optimal BP neural network prediction model is output;
s5: inputting all production data serving as initial parameters into a GA-BP neural network prediction model, completing learning, parameter adjustment, threshold matching and result judgment of the GA-BP neural network prediction model to obtain an optimal BP neural network prediction model, and outputting the optimal GA-BP neural network prediction model;
s6: using the optimal BP neural network prediction model to predict the results of the past production data to obtain a first prediction result containing a process parameter combination, using the optimal GA-BP neural network prediction model to predict the results of the past production data to obtain a second prediction result containing the process parameter combination,
s7: setting a final judgment module, feeding back the first prediction result and the second prediction result to the final judgment module by the data processing module, obtaining an optimal process parameter combination by a processor in the final judgment module according to the first prediction result and the second prediction result, and guiding the production process of a product by using the optimal process parameter combination.
Preferably, the data monitoring module in step S1 establishes an injection molding theoretical model, the injection molding theoretical model includes a plastic melt flow model and a warpage deformation model, the flow model is mainly used for analyzing the cause of warpage deformation to obtain corresponding production parameter data, the warpage deformation model is mainly used for analyzing the multivariable and nonlinear melt flow change process in the injection molding process to obtain corresponding production parameter data, computer-aided modeling in the engineering building of a pouring system and a cooling system is performed on the thin shell plastic part according to the shape and material properties of the thin shell plastic part and by combining with data recommended by a Moldflow database, and preliminary setting of molding process parameters is performed according to the recommendation, so as to establish an injection molding numerical simulation system, implement preliminary warpage deformation and weld mark defect analysis on the thin shell plastic part, and obtain production defective data and qualified product data of the thin shell plastic part.
Preferably, all the production data in the step S1 include production parameter data, defective product data, and qualified product data, the production parameter data specifically includes filling time, melt temperature, mold temperature, holding pressure time, and cooling time, the defective product data and the qualified product data are determined by data corresponding to the product to be produced, and the purposes of rapidly finding defects and providing a solution are achieved through a computer aided engineering technology in engineering design, shortening the mold manufacturing period, and accelerating new product development and rapid marketing.
Preferably, the Moldflow database is obtained based on a finite element numerical analysis software Moldflow, the Moldflow software is a mold flow analysis, the mold flow analysis refers to that data simulation software is used, simulation of injection molding is completed through a computer, the injection molding process of a mold is simulated, a corresponding data result is obtained, and feasibility of a production scheme of the mold is evaluated through the corresponding result.
Preferably, the BP neural network prediction model in step S3 optimizes the weight and threshold of the BP neural network training mainly through a genetic algorithm, and first performs coding to generate an initial population, trains the neural network with the genetic algorithm, and adopts real number coding; and then, executing genetic operation, calculating the fitness value of each chromosome in the current population, finding out the individual with the current optimal fitness value, repeating iteration until the condition is met, if the condition is not met, taking the specified maximum genetic algebra as a termination calculation criterion, finally obtaining the initial weight and the threshold of the BP neural network, and obtaining a group of complete initial weight and threshold with the minimum error of the BP neural network prediction model through genetic operation.
Preferably, in the real number encoding process, a real number is directly used as a locus of a chromosome, so that the length of the chromosome is greatly shortened, the complexity of encoding and decoding is avoided, the genetic operation is simplified, and an encoding string consists of four parts: the method comprises the following steps that a hidden layer is connected with an input layer weight, an output layer is connected with a hidden layer weight, a hidden layer threshold and an output layer threshold, and the specific method comprises the following steps: the weight values and threshold values of the network are cascaded according to a certain sequence to form a real number array as a chromosome of a genetic algorithm, genetic operation is carried out in the chromosome group, then an evaluation function is carried out, the genetic algorithm is based on a fitness function in evolutionary search, the fitness value of each chromosome in the population is used for searching, the probability that an individual with higher fitness is inherited to the next generation is higher, the probability that an individual with lower fitness is inherited to the next generation is relatively lower, and the evaluation standard of the BP neural network is that the MSE is better and smaller.
Preferably, in the process of executing the genetic operation, the selection operation adopts a sorting selection method, the selection operation is arranged from small to large according to the size of each individual fitness value, the serial number corresponding to the individual with the minimum fitness value is 1, the serial number corresponding to the individual with the maximum fitness value is M, then the selection probability of the individual is calculated according to the fitness proportion selection method according to the size of the relative fitness value of each individual, the cross operation adopts single-point cross, and the mutation operation adopts uniform mutation.
Compared with the prior art, the invention has the following beneficial effects:
the invention mainly researches the most common product defects in the production process of the thin-shell plastic parts, realizes the purposes of quickly finding the defects and providing a solution through the CAE technology, shortening the manufacturing period of the die and accelerating the development and the quick marketing of new products.
According to the invention, firstly, a data set obtained by Moldflow software is used as input data and output data of a BP neural network and a GA-BP network, the shrinkage rate of a plastic part can be predicted through the prediction capability of a BP neural network model, the prediction results of the GA-BP network model are compared, and the simulation prediction results of two network calculation models are compared, so that the optimal process parameter combination of the injection molding process can be obtained, the prediction result of a neural network algorithm can be further optimized through the coupling of a genetic algorithm and the neural network algorithm, and a prediction model with higher accuracy and stability is obtained.
According to the invention, two prediction results are obtained by establishing two neural network prediction models, product production parameters are finally determined according to the two prediction results and by combining manual judgment, and product production is guided according to the product production parameters in the predicted results, so that the data analysis result is more accurate, the waste of human resources is saved, and errors are avoided in the product production process.
Drawings
FIG. 1 is a flow chart of a prediction method for product production based on a neural network model of the present invention;
FIG. 2 is a system diagram of a prediction method for product production based on neural network model according to the present invention;
FIG. 3 is a flow chart of a BP neural network prediction model in a prediction method for product production based on a neural network model of the present invention.
Detailed Description
In order to make the technical means, the creation characteristics, the achievement purposes and the effects of the invention easy to understand, the invention is further described with the specific embodiments.
In the description of the present invention, it should be noted that the terms "upper", "lower", "inner", "outer", "front", "rear", "both ends", "one end", "the other end", and the like indicate orientations or positional relationships based on those shown in the drawings, and are only for convenience of description and simplicity of description, but do not indicate or imply that the referred device or element must have a specific orientation, be constructed in a specific orientation, and be operated, and thus, should not be construed as limiting the present invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In the description of the present invention, it is to be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "disposed," "connected," and the like are to be construed broadly, such as "connected," which may be fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
Referring to fig. 1-3, the present invention is a prediction method for product production based on neural network model, comprising the following steps:
s1: setting a data monitoring module, wherein the data monitoring module acquires all production data in the injection molding production process of the thin-shell plastic part in real time;
s2: the data monitoring system is characterized in that a data transmission module and a data processing module are arranged, the data transmission module is mainly used for sending all production data to the data processing module, the data processing module is mainly used for processing all production data, the data monitoring module is connected with the data processing module through the data transmission module, the data transmission module is specifically a communication network, data transmission is achieved through the communication network, and meanwhile stability of data transmission is guaranteed;
s3: a double neural network prediction unit is established in the data processing module, and a BP neural network prediction model and a GA-BP neural network prediction model are established in the double neural network prediction unit;
s4: inputting all production data serving as initial parameters into a BP neural network prediction model, completing learning, parameter adjustment, threshold matching and result judgment of the BP neural network prediction model to obtain an optimal BP neural network prediction model, and outputting the optimal BP neural network prediction model;
s5: inputting all production data serving as initial parameters into a GA-BP neural network prediction model, completing learning, parameter adjustment, threshold value matching and result judgment of the GA-BP neural network prediction model to obtain an optimal BP neural network prediction model, and outputting the optimal GA-BP neural network prediction model;
s6: using the optimal BP neural network prediction model to predict the results of the past production data to obtain a first prediction result containing a process parameter combination, using the optimal GA-BP neural network prediction model to predict the results of the past production data to obtain a second prediction result containing the process parameter combination,
s7: setting a final judgment module, feeding back the first prediction result and the second prediction result to the final judgment module by the data processing module, obtaining an optimal process parameter combination by the processing personnel in the final judgment module according to the first prediction result and the second prediction result, and guiding the production process of the product by using the optimal process parameter combination.
The method comprises the following steps that in the step S1, an injection molding theoretical model is established in a data monitoring module and comprises a plastic melt flow model and a warping deformation model, the flow model is mainly used for analyzing reasons causing warping deformation to obtain corresponding production parameter data, the warping deformation model is mainly used for analyzing a multivariable and nonlinear melt flow change process in the injection molding process to obtain corresponding production parameter data, computer-aided modeling in engineering buildings of a pouring system and a cooling system is carried out on the thin-shell plastic part according to the shape and material properties of the thin-shell plastic part and data recommended by a Moldflow database, and molding process parameters are preliminarily set according to the recommendation, so that an injection molding numerical simulation system is established, preliminary warping deformation and welding mark defect analysis of the thin-shell plastic part is realized, and production defective data and qualified product data of the thin-shell are obtained.
All production data in the step S1 comprise production parameter data, defective product data and qualified product data, the production parameter data specifically comprise filling time, melt temperature, mold temperature, pressure maintaining pressure, pressure maintaining time and cooling time, the defective product data and the qualified product data are determined by data corresponding to a product to be produced, defects are quickly found and a solution is given through a computer aided engineering technology in engineering design, the mold manufacturing period is shortened, and the purposes of new product development and quick marketing are accelerated.
The method comprises the steps of using a data set obtained by the Moldflow software as input data and output data of a BP neural network and a GA-BP network, predicting the shrinkage rate of a plastic part through the prediction capability of the BP neural network model, comparing the prediction results of the GA-BP network model, and calculating the simulation prediction results of the models through comparing two networks, so that the optimal process parameter combination of the injection molding process can be obtained, and the prediction result of the neural network algorithm can be further optimized through the coupling of a genetic algorithm and the neural network algorithm, so that a prediction model with higher accuracy and stability can be obtained.
The BP neural network prediction model in the step S3 optimizes the weight and the threshold value of BP neural network training mainly through a genetic algorithm, firstly carries out coding to generate an initial population, trains the neural network through the genetic algorithm, and adopts real number coding; and then executing genetic operation, then calculating the fitness value of each chromosome in the current population, finding out the individual of the current optimal fitness value, iterating repeatedly until the condition is met, if the condition is not met, taking the specified maximum genetic algebra as a termination calculation criterion, finally obtaining the initial weight and the threshold of the BP neural network, and obtaining a group of complete initial weights and thresholds with the minimum error of the BP neural network prediction model through genetic operation.
Wherein, in the real number coding process of adoption, directly as a gene locus of a chromosome with a real number, make the length of chromosome shorten greatly, removed the loaded down with trivial details of encoding and decoding from for genetic operation simplifies, and the encoding string comprises four parts: the method comprises the following steps that a hidden layer is connected with an input layer through a weight, an output layer is connected with a hidden layer through a weight, a hidden layer threshold and an output layer threshold, and the specific method comprises the following steps: the weight values and threshold values of the network are cascaded according to a certain sequence to form a real number array as a chromosome of a genetic algorithm, genetic operation is carried out in the chromosome group, then an evaluation function is carried out, the genetic algorithm is based on a fitness function in evolutionary search, the fitness value of each chromosome in the population is used for searching, the probability that an individual with higher fitness is inherited to the next generation is higher, the probability that an individual with lower fitness is inherited to the next generation is relatively lower, and the evaluation standard of the BP neural network is that the MSE is better and smaller.
In the process of executing the genetic operation, the selection operation adopts a sorting selection method, the selection operation is arranged from small to large according to the size of each individual fitness value, the serial number corresponding to the individual with the minimum fitness value is 1, the serial number corresponding to the individual with the maximum fitness value is M, then the selection probability of the individual is calculated according to the fitness proportion selection method according to the size of the relative fitness value of each individual, the cross operation adopts single-point cross, and the mutation operation adopts uniform mutation.
In actual use, mold flow analysis of a novel mold of 50 pairs/year is realized, the manufacturing and mold adjusting and testing period is shortened by about 180 pairs/year, the research and development cost and the production cost are reduced by 140 thousands/year, proper gate positions, a pouring system and a cooling system can be obtained by analysis through CAE simulation prediction software and Moldflow simulation prediction, and after proper materials are selected, the injection molding process is guided through Moldflow software by designing specific process parameter combinations.
The foregoing shows and describes the general principles and features of the present invention, together with the advantages thereof. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (7)
1. A prediction method for product production based on a neural network model is characterized in that: the method comprises the following steps:
s1: setting a data monitoring module, wherein the data monitoring module acquires all production data in the injection molding production process of the thin-shell plastic part in real time;
s2: the method comprises the steps that a data transmission module and a data processing module are set, wherein the data transmission module is mainly used for sending all production data to the data processing module, and the data processing module is mainly used for processing all production data;
s3: establishing a dual neural network prediction unit in the data processing module, and establishing a BP neural network prediction model and a GA-BP neural network prediction model in the dual neural network prediction unit;
s4: inputting all production data serving as initial parameters into a BP neural network prediction model, completing learning, parameter adjustment, threshold matching and result judgment of the BP neural network prediction model to obtain an optimal BP neural network prediction model, and outputting the optimal BP neural network prediction model;
s5: inputting all production data serving as initial parameters into a GA-BP neural network prediction model, completing learning, parameter adjustment, threshold matching and result judgment of the GA-BP neural network prediction model to obtain an optimal BP neural network prediction model, and outputting the optimal GA-BP neural network prediction model;
s6: the method comprises the steps that an optimal BP neural network prediction model is used for conducting result prediction on past production data to obtain a first prediction result containing a process parameter combination, and an optimal GA-BP neural network prediction model is used for conducting result prediction on the past production data to obtain a second prediction result containing the process parameter combination;
s7: and setting a final judgment module, feeding the first prediction result and the second prediction result back to the final judgment module by the data processing module, obtaining an optimal process parameter combination by the processing personnel in the final judgment module by combining the first prediction result and the second prediction result, and guiding the production process of the product by using the optimal process parameter combination.
2. The neural network model-based product production prediction method of claim 1, wherein: the method comprises the following steps that an injection molding theoretical model is established in a data monitoring module in the step S1, the injection molding theoretical model comprises a plastic melt flow model and a warping deformation model, the flow model is mainly used for analyzing reasons causing warping deformation to obtain corresponding production parameter data, the warping deformation model is mainly used for analyzing a multivariable and nonlinear melt flow change process in the injection molding process to obtain corresponding production parameter data, computer-aided modeling in engineering buildings of a pouring system and a cooling system is carried out on the thin shell plastic part according to the shape and material properties of the thin shell plastic part and data recommended by a Moldflow database, and molding process parameters are preliminarily set according to the recommendation, so that an injection molding numerical simulation system is established, preliminary warping deformation and welding mark defect analysis of the thin shell plastic part are achieved, and production defective goods data and qualified product data of the thin shell are obtained.
3. The prediction method for product production based on the neural network model according to claim 2, wherein: all production data in the step S1 comprise production parameter data, defective product data and qualified product data, the production parameter data specifically comprise filling time, melt temperature, mold temperature, pressure maintaining pressure, pressure maintaining time and cooling time, and the defective product data and the qualified product data are determined by data corresponding to specific products to be produced.
4. The neural network model-based product production prediction method as set forth in claim 2, wherein: the Moldflow database is obtained based on finite element numerical analysis software Moldflow, the Moldflow software is mold flow analysis, the mold flow analysis refers to the steps of using data simulation software, completing simulation of injection molding through a computer, simulating the injection molding process of a mold, obtaining corresponding data results, and evaluating the feasibility of the production scheme of the mold through the corresponding results.
5. The neural network model-based product production prediction method as set forth in claim 1, wherein: the BP neural network prediction model in the step S3 is mainly used for optimizing the weight and the threshold value of BP neural network training through a genetic algorithm, firstly coding is carried out to generate an initial population, the neural network is trained through the genetic algorithm, and real number coding is adopted; and then, executing genetic operation, calculating the fitness value of each chromosome in the current population, finding out the individual with the current optimal fitness value, repeating iteration until the condition is met, if the condition is not met, taking the specified maximum genetic algebra as a termination calculation criterion, finally obtaining the initial weight and the threshold of the BP neural network, and obtaining a group of complete initial weight and threshold with the minimum error of the BP neural network prediction model through genetic operation.
6. The prediction method for product production based on the neural network model according to claim 5, wherein: in the real number encoding process, a real number is directly used as a gene locus of a chromosome, and an encoding string consists of four parts: the method comprises the following steps that a hidden layer is connected with an input layer through a weight, an output layer is connected with a hidden layer through a weight, a hidden layer threshold and an output layer threshold, and the specific method comprises the following steps: the weight values and the threshold values of the network are cascaded according to a certain sequence to form a real number array as a chromosome of a genetic algorithm, genetic operation is carried out in the chromosome group, then an evaluation function is carried out, the genetic algorithm is based on a fitness function in evolutionary search, and the fitness value of each chromosome in the population is utilized for search.
7. The prediction method for product production based on the neural network model according to claim 5, wherein: in the process of executing the genetic operation, the selection operation adopts a sorting selection method, the selection operation is arranged from small to large according to the size of each individual fitness value, the serial number corresponding to the individual with the minimum fitness value is 1, the serial number corresponding to the individual with the maximum fitness value is M, then the selection probability of the individual is calculated according to the fitness proportion selection method according to the size of the relative fitness value of each individual, the cross operation adopts single-point cross, and the mutation operation adopts uniform mutation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211072340.6A CN115409270A (en) | 2022-09-02 | 2022-09-02 | Neural network model based product production prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211072340.6A CN115409270A (en) | 2022-09-02 | 2022-09-02 | Neural network model based product production prediction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115409270A true CN115409270A (en) | 2022-11-29 |
Family
ID=84164170
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211072340.6A Pending CN115409270A (en) | 2022-09-02 | 2022-09-02 | Neural network model based product production prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115409270A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115879826A (en) * | 2023-02-20 | 2023-03-31 | 深圳普菲特信息科技股份有限公司 | Fine chemical process quality inspection method, system and medium based on big data |
-
2022
- 2022-09-02 CN CN202211072340.6A patent/CN115409270A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115879826A (en) * | 2023-02-20 | 2023-03-31 | 深圳普菲特信息科技股份有限公司 | Fine chemical process quality inspection method, system and medium based on big data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109657805A (en) | Hyper parameter determines method, apparatus, electronic equipment and computer-readable medium | |
| CN107193876A (en) | A kind of missing data complementing method based on arest neighbors KNN algorithms | |
| CN112257963B (en) | Defect prediction method and device based on spaceflight software defect data distribution outlier | |
| CN109858714B (en) | Tobacco shred quality inspection index prediction method, device and system based on improved neural network | |
| CN115409270A (en) | Neural network model based product production prediction method | |
| AlKaabneh et al. | A combined analytical hierarchical process (AHP) and Taguchi experimental design (TED) for plastic injection molding process settings | |
| CN110232461A (en) | More interconnection vector machine water quality prediction methods based on quantum genetic algorithm optimization | |
| CN108399470B (en) | Indoor PM2.5 prediction method based on multi-example genetic neural network | |
| CN111428329B (en) | Model-based machine learning system | |
| CN111898867B (en) | Airplane final assembly production line productivity prediction method based on deep neural network | |
| CN113033786B (en) | Fault diagnosis model construction method and device based on time convolution network | |
| CN111785326B (en) | Gene expression profile prediction method after drug action based on generation of antagonism network | |
| CN117274259A (en) | Keyboard production equipment management method and system based on Internet of things | |
| CN109117529A (en) | Resistance model for prediction physical parameter extracting method | |
| CN111222762A (en) | Solar cell panel coating process state monitoring and quality control system | |
| CN106682938A (en) | Big-data precision marketing model establishing method and device | |
| CN112882934A (en) | Test analysis method and system based on defect growth | |
| CN117334271A (en) | Method for generating molecules based on specified attributes | |
| CN111571913A (en) | Method for predicting defects of bathroom plastic parts | |
| CN116189827A (en) | Object-oriented digital twin space structure model generation and life-span management method | |
| Ma et al. | Intelligent recommendation system of the injection molding process parameters based on CAE simulation, process window, and machine learning | |
| CN115222114A (en) | Automobile part assembly index value prediction method, terminal device and storage medium | |
| CN115408842A (en) | Method for analyzing similarity of characteristics of injection molding simulation process | |
| CN111210877B (en) | Method and device for deducing physical parameters | |
| CN114881289B (en) | Digital twinning-based fusion-casting charge forming process parameter optimization method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |