Search

JP-2026075205-A - Molding condition prediction device, molding condition prediction method, and computer program

JP2026075205AJP 2026075205 AJP2026075205 AJP 2026075205AJP-2026075205-A

Abstract

[Problem] To provide a technology that shortens the time required to search for molding conditions for a molded product in a molding condition prediction device. [Solution] The molding condition prediction device includes: a storage unit that stores a first prediction model which includes information on the relationship between a second molding condition and a feature quantity when the value corresponding to a first molding condition, which is one of the molding conditions for a molded product, is a1; a second prediction model which includes information on the relationship between a second molding condition and a feature quantity when the value corresponding to the first molding condition is a2; a prediction model generation unit which uses the first prediction model and the second prediction model to generate a new prediction model which includes information on the relationship between a second molding condition and a feature quantity when the value corresponding to the first molding condition is a3; a molding condition prediction unit which uses the new prediction model to predict the value corresponding to the second molding condition when the value corresponding to the first molding condition is a3; and an output unit which outputs the result predicted by the molding condition prediction unit. [Selection Diagram] Figure 3

Inventors

  • 藤村 南
  • 太田 英一
  • 佐藤 康元
  • 三国 敦
  • 稲森 茂
  • 上野 紀幸

Assignees

  • 株式会社豊田中央研究所
  • トヨタ自動車株式会社

Dates

Publication Date
20260508
Application Date
20241022

Claims (9)

  1. A molding condition prediction device, A storage unit for storing a predictive model that includes information relating to the relationship between the molding conditions of a molded product and a feature quantity relating to the molded product, the storage unit for storing a first predictive model that includes information relating to the relationship between a second molding condition and the feature quantity when the value corresponding to a first molding condition, which is one of the molding conditions, is a1, and a second predictive model that includes information relating to the relationship between the second molding condition and the feature quantity when the value corresponding to the first molding condition is a2, which is different from a1, A prediction model generation unit generates a new prediction model that uses the first prediction model and the second prediction model to generate a new prediction model that includes information regarding the relationship between the second molding condition and the feature quantity when the value corresponding to the first molding condition is a3 which is different from a1 and a2, A molding condition prediction unit that uses the new prediction model to predict the value corresponding to the second molding condition when the value corresponding to the first molding condition is a3, The system includes an output unit that outputs the results predicted by the molding condition prediction unit. Molding condition prediction device.
  2. The molding condition prediction device described in claim 1 further, The system includes an information acquisition unit that acquires information regarding a3 as a value corresponding to the first molding condition that is set in advance for the target molded product which is the target of the molding condition prediction. Molding condition prediction device.
  3. A molding condition prediction device according to claim 2, The aforementioned prediction model generation unit, In the first prediction model, the first feature quantity is calculated when the value corresponding to the second molding condition is b. In the second prediction model, when the value corresponding to the second molding condition is b, a second feature quantity is calculated. Using the first feature quantity and the second feature quantity, the target feature quantity for the target molded product is calculated when the value corresponding to the first molding condition is a3 and the value corresponding to the second molding condition is b. Molding condition prediction device.
  4. A molding condition prediction device according to claim 3, The prediction model generation unit calculates the target feature by applying linear interpolation to the relationship between the first feature and the second feature. Molding condition prediction device.
  5. A molding condition prediction device according to claim 2 or claim 3, The information acquisition unit acquires information regarding the setting range of the second molding conditions when molding the target molded product, The aforementioned storage unit is A prediction model including information relating to the relationship between the second molding conditions set within the range including the aforementioned setting range and the feature quantities is stored as the first prediction model and the second prediction model, The prediction model generation unit generates a new prediction model that includes information regarding the relationship between the second molding conditions and the target feature quantity within the range including the set range. Molding condition prediction device.
  6. The molding condition prediction device according to claim 1 or claim 2 further, An evaluation unit that evaluates the results predicted by the molding condition prediction unit, The system includes a prediction model modification unit that modifies the new prediction model using the evaluation results from the evaluation unit, Molding condition prediction device.
  7. A molding condition prediction device according to claim 6, The molding condition prediction unit uses the new prediction model modified by the prediction model modification unit to predict a new value corresponding to the second molding condition when the value corresponding to the first molding condition is a3. The output unit outputs the new value predicted by the molding condition prediction unit. Molding condition prediction device.
  8. A molding condition prediction method for predicting the molding conditions of a molded product using a molding condition prediction device, A storage step for storing a predictive model that includes information relating to the relationship between the molding conditions of a molded product and a feature quantity relating to the molded product, comprising: a first predictive model that includes information relating to the relationship between a second molding condition and the feature quantity when the value corresponding to a first molding condition, which is one of the molding conditions, is a1; and a second predictive model that includes information relating to the relationship between the second molding condition and the feature quantity when the value corresponding to the first molding condition is a2, which is different from a1. A prediction model generation step of generating a new prediction model that includes information regarding the relationship between the second molding condition and the feature quantity when the value corresponding to the first molding condition is a3 which is different from a1 and a2, using the first prediction model and the second prediction model, A molding condition prediction step, in which the value corresponding to the second molding condition is predicted when the value corresponding to the first molding condition is a3 using the new prediction model, The system includes an output step that outputs the predicted results in the molding condition prediction step. A method for predicting molding conditions.
  9. A computer program that causes a computer to predict the molding conditions for a molded product, A memory function for storing a predictive model that includes information relating to the relationship between the molding conditions of a molded product and a feature quantity relating to the molded product, the memory function for storing a first predictive model that includes information relating to the relationship between a second molding condition and the feature quantity when the value corresponding to a first molding condition, which is one of the molding conditions, is a1, and a second predictive model that includes information relating to the relationship between the second molding condition and the feature quantity when the value corresponding to the first molding condition is a2, which is different from a1, A prediction model generation function that generates a new prediction model that includes information regarding the relationship between the second molding condition and the feature quantity when the value corresponding to the first molding condition is a3 which is different from a1 and a2, using the first prediction model and the second prediction model, A molding condition prediction function that uses the new prediction model to predict the value corresponding to the second molding condition when the value corresponding to the first molding condition is a3, An output function that outputs the results predicted by the molding condition prediction function, and a function that causes the computer to execute Computer program.

Description

This invention relates to a molding condition prediction device, a molding condition prediction method, and a computer program. Conventionally, molding condition prediction devices that use predictive models to predict the molding conditions for forming pre-set molded products are known (for example, Patent Document 1). Japanese Patent Publication No. 2019-28949 This is a schematic diagram showing the general configuration of the molding system of the first embodiment.This figure illustrates a molded product manufactured by the molding system of the first embodiment.This is a block diagram illustrating the configuration of the molding condition prediction device according to the first embodiment.This is a diagram illustrating the molding conditions for molded products.This is a diagram illustrating the characteristic features of molded products.This diagram illustrates the predictive model stored in the memory unit.This diagram illustrates the method for generating a new prediction model using the prediction model generation unit.This diagram illustrates the method for predicting molding conditions using the molding condition prediction unit.This figure illustrates the molding condition prediction method of the first embodiment.This diagram illustrates the molding method for molded products in evaluation tests.This diagram illustrates the predicted shape and movement of the tool in the evaluation test.This is a diagram illustrating the test results of the evaluation test.This is a block diagram illustrating the configuration of the molding condition prediction device according to the second embodiment. <First Embodiment> Figure 1 is a schematic diagram showing the general configuration of the molding system 100 of the first embodiment. The molding system 100 is a system for molding a molded product from a member to be molded, and comprises a molding apparatus 10 and a molding condition prediction device 1. In the molding system 100 of this embodiment, the molding apparatus 10 molds the member to be molded using the molding conditions for the molded product predicted by the molding condition prediction device 1, thereby producing a molded product having a preset shape. This makes it possible to shorten the time required to search for molding conditions for molding a molded product. The molding apparatus 10 is a device that actually molds a molded product Pm using the molding conditions predicted by the molding condition prediction device 1. The molding apparatus 10 comprises a molding unit 11 and a control unit 12. The molding unit 11 includes a plurality of tools T1, T2, T3 that contact the material to be molded product Pm and deform the material, and actuators Ac1, Ac2, Ac3 for driving the plurality of tools T1, T2, T3. The control unit 12 is electrically connected to the molding condition prediction device 1 and controls the driving of the actuators Ac1, Ac2, Ac3 in the molding unit 11 using the molding conditions output by the molding condition prediction device 1. As a result, the molding apparatus 10 can mold a molded product Pm having a preset shape. Figure 2 illustrates the molded product Pm manufactured by the molding system 100 of this embodiment. Figure 2 shows perspective views of the sheet material Pb supplied to the molding apparatus 10 and the molded product Pm manufactured by molding the sheet material Pb. The sheet material Pb is a plate-shaped member having a roughly fan-like shape. The molded product Pm, which is the target molded product for which molding conditions are predicted, is a so-called hat-shaped member, comprising a top plate portion Pm1, a flange portion Pm2, a wall portion Pm3 connecting the top plate portion Pm1 and the flange portion Pm2, and a connecting portion Pm4 connecting the top plate portion Pm1 and the wall portion Pm3. In the molded product Pm, a step is formed by the wall portion Pm3. The connecting portion Pm4 is formed in a curved shape, as shown in Figure 2. The molding condition prediction device 1 uses feature quantities related to the molded product Pm to predict and output molding conditions necessary for the molded product Pm to take on a preset shape. In this embodiment, the molding condition prediction device 1 is a personal computer (hereinafter simply referred to as "PC") 20. The PC 20 includes an input unit 20a, a storage medium 20b, a CPU (Central Processing Unit) 20c, and an output unit 20d. The input unit 20a is at least one of the following: a keyboard, a mouse, an interface to an external recording device such as a USB memory or flash memory, or a receiver capable of receiving signals from the outside. The input unit 20a accepts inputs regarding the type and upper and lower limits of feature quantities related to the molded product Pm, which is the target of molding condition prediction, and information regarding the types and controllable range of molding conditions that can be controlled by the molding apparatus 10. The storage medium 20b is a