Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Method for inverting simulation parameters through machine learning neural network

A neural network and machine learning technology, applied in the field of inversion of simulation parameters through machine learning neural network, can solve problems such as simulation result deviation, error numerical model, error programming, etc., to reduce simulation errors and minimize differences.

Pending Publication Date: 2020-09-29
深圳同奈信息科技有限公司 +1
View PDF0 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, these prediction results will still be different from the reality in the injection molding process. For example, the difference between the simulation environment and the real environment will cause errors, numerical model errors in the simulation software, or programming errors, which may make the simulation There are deviations in the results, and there is a problem that the degree of matching between simulation and reality is not enough

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Method for inverting simulation parameters through machine learning neural network
  • Method for inverting simulation parameters through machine learning neural network
  • Method for inverting simulation parameters through machine learning neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0039] see Figure 1 to Figure 8 , a method for inverting simulation parameters through a machine learning neural network, comprising the following steps:

[0040] S1, establish a machine learning neural network, including an input layer and an output layer, the test conditions of the input layer and the output layer are the same as those of the physical model;

[0041] S2, use a simple physical model as the basis for simulation prediction, simulating multiple possible working conditions of the injection molding process of the designed plastic parts; this embodiment simulates 18 working conditions of its plastic parts, and the input data is as follows image 3 shown;

[0042] S3. In the machine learning neural network and physical model, input the same original parameters according to the injection molding process, and use the physical model as the learning object to invert and adjust the operation of the machine learning neural network to match the third weight set W3 , min...

Embodiment 2

[0054] This embodiment specifically describes the training of the machine learning neural network model. The difference from the above embodiment is that the rheological model parameters of the recycled polypropylene plastic corresponding to the original plastic in Example 1 are reversed. The rheological model parameters in this embodiment Specifically, model parameters include at least one set of shear viscosity model parameters and a set of additional model parameters for calculating pressure loss in case of constricted flow. This additional model requires that a pressure transducer be placed before and after the systolic flow to measure the pressure loss of the systolic flow. Additional model parameters can be better measured by designing and using at least two or more different constricted flows.

[0055] On measurement data, two forms are acceptable:

[0056] 1) A constant value at a certain moment, for example, the moment corresponding to the peak value measured by one ...

Embodiment 3

[0081] Propose a spiral mold in this embodiment, the physical model of this embodiment is injection mold flow analysis software, also can be the physical model of any injection mold flow analysis in other embodiments, use three melt temperature conditions equally, be respectively 220 °C, 240 °C, and 260 °C have corresponding speed and pressure values ​​at each melt temperature, and the specific values ​​are in Figure 10 listed in . The material used for the experiments was polypropylene Stamylan PHC 31, as Figure 11 Shown is a mold flow analysis of the spiral plastic part used, a simulation of the spiral mold was performed to provide a predictive equation based on a second order viscosity model as follows:

[0082]

[0083] Wherein η is viscosity (Pa.s); A i model coefficients; Shear rate (1 / s); T temperature (°C).

[0084] Similarly, in this embodiment, the second-order viscosity model is used as the training model of the neural network model of the present inventio...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a method for inverting simulation parameters through a machine learning neural network. The method comprises the following steps: establishing the machine learning neural network; and taking injection molding simulation software based on the physical model as a learning object, and adjusting the weight of the neural network by using an inversion algorithm in a machine learning algorithm to enable the neural network to use the same input parameter to calculate a corresponding result calculated by the injection molding simulation software based on the physical model. Andthen, according to the difference between the measured value in the injection molding system quantified by the sensor and the predicted value of the neural network model learned by the training, the input parameters which should be used in the injection molding simulation software are inverted, the most important parameters being, but not limited to, rheological model coefficients. And the input parameters subjected to machine learning inversion adjustment are introduced to the injection molding simulation software, so that the difference between a simulation prediction value and an actual measurement value can be reduced, and digital twinning of a simulation injection molding process is realized so as to facilitate intelligent control.

Description

technical field [0001] The invention relates to the field of intelligent control of physical model simulation, in particular to a method for inverting simulation parameters through a machine learning neural network. Background technique [0002] After the design of plastic parts is completed, it is manufactured by injection molding process. Firstly, certain computer simulation calculations are carried out on various working conditions, so that the predicted value of process conditions can be obtained by mold flow analysis and used for trial mold adjustment. After at least one trial mold adjustment Obtaining sensor measurements installed in the mold, and there are differences between these measurements and the mold flow software predictions. It is the existence of these differences that may prevent the establishment of a reliable intelligent control system. [0003] Usually, simulation software can quantitatively predict various possible results, provide a set of good referen...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06F30/27G06F30/17G06N3/04G06N3/08G06F113/22
CPCG06F30/27G06F30/17G06N3/084G06F2113/22G06N3/045
Inventor 金小石
Owner 深圳同奈信息科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products