Method and device for predicting defects
A method and device for predicting a defect. The method includes determining a sequence between a plurality of sub-models by modeling a production process into the plurality of sub-models, mapping production process data into each of the plurality of sub-models, determining, by a corresponding sub-model, output data comprising defect information on a potential defect occurring in a corresponding step, for each of the plurality of sub-models, predicting information associated with a defect in the production process based on the output data corresponding to each of the plurality of sub-models, and inputting the output data of each of the sub-models to a subsequent sub-model of the corresponding sub-model, based on the sequence.
1. A processor-implemented defect prediction method, comprising:
determining a sequence, being a sequential order of execution of a plurality of sub-models, by modeling a production process into the plurality of sub-models;
mapping production process data into each of the plurality of sub-models;
determining, by a corresponding sub-model, output data comprising defect information on a potential defect occurring in a corresponding step based on predicted third data of the corresponding sub-model and measured third data, for each of the plurality of sub-models;
predicting information associated with a defect in the production process based on the output data corresponding to each of the plurality of sub-models; and
inputting the output data of each of the sub-models to a subsequent sub-model of the corresponding sub-model, based on the sequence.
2. The processor-implemented defect prediction method of claim 1 , further comprising:
obtaining the production process data.
3. The processor-implemented defect prediction method of claim 1 , wherein the production process data comprises any one of or any combination of:
first data measured as the production process is performed;
second data processed based on the first data; and
the measured third data measured to determine whether a defect is present.
4. The processor-implemented defect prediction method of claim 1 , wherein the determining of the output data comprises:
receiving previous output data of a previous sub-model of the corresponding sub-model; and
determining the output data of the corresponding sub-model based on the previous output data and the production process data of the corresponding sub-model.
5. The processor-implemented defect prediction method of claim 4 , wherein the determining of the output data of the corresponding sub-model comprises:
predicting the predicted third data of the corresponding sub-model based on the previous output data, first data of the corresponding sub-model, and second data of the corresponding sub-model; and
determining final third data based on the predicted third data and the measured third data.
6. The processor-implemented defect prediction method of claim 5 , wherein the determining of the final third data comprises:
determining the measured third data to be the final third data in response to the measured third data being present; and
determining the predicted third data to be the final third data in response to the measured third data being absent.
7. The processor-implemented defect prediction method of claim 5 , wherein the determining of the output data of the corresponding sub-model comprises:
generating latent space information of the corresponding sub-model based on the previous output data, the first data of the corresponding sub-model, and the second data of the corresponding sub-model; and
determining the output data based on any one of or any combination of the first data of the corresponding sub-model and the second data of the corresponding sub-model, the final third data, and the latent space information.
8. The processor-implemented defect prediction method of claim 1 , wherein the mapping of the production process data into each of the plurality of sub-models comprises dividing the production process data based on steps.
9. The processor-implemented defect prediction method of claim 1 , wherein the determining of the sequence comprises determining the sequence between the plurality of sub-models based on any one of or any combination of a temporal order and a physical correlation of the production process.
10. A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the defect prediction method of claim 1 .
11. A defect prediction device, comprising:
a processor configured to:
determine a sequence, being a sequential order of execution of a plurality of sub-models, by modeling a production process into the plurality of sub-models,
map production process data into each of the plurality of sub-models, determine, by a corresponding sub-model, output data comprising defect information on a potential defect occurring in a corresponding step based on predicted third data of the corresponding sub-model and measured third data, for each of the plurality of sub-models, and
predict information associated with a defect in the production process based on the output data corresponding to each of the plurality of sub-models,
wherein the output data of each of the sub-models is input to a subsequent sub-model of the corresponding sub-model, based on the sequence.
12. The defect prediction device of claim 11 , wherein the processor is configured to obtain the production process data.
13. The defect prediction device of claim 11 , wherein the production process data comprises any one of or any combination of:
first data measured as the production process is performed;
second data processed based on the first data; and
the measured third data measured to determine whether a defect is present.
14. The defect prediction device of claim 11 , wherein the processor is configured to:
receive previous output data of a previous sub-model of the corresponding sub-model; and
determine the output data of the corresponding sub-model based on the previous output data and the production process data of the corresponding sub-model.
15. The defect prediction device of claim 14 , wherein the processor is configured to:
predict the predicted third data of the corresponding sub-model based on the previous output data, the first data of the corresponding sub-model, and the second data of the corresponding sub-model; and
determine final third data based on the predicted third data and the measured third data.
16. The defect prediction device of claim 15 , wherein the processor is configured to:
determine the measured third data to be the final third data in response to the measured third data being present; and
determine the predicted third data to be the final third data in response to the measured third data being absent.
17. The defect prediction device of claim 15 , wherein the processor is configured to:
generate latent space information of the corresponding sub-model based on the previous output data, the first data of the corresponding sub-model, and the second data of the corresponding sub-model; and
determine the output data based on any one of or any combination of the first data of the corresponding sub-model and the second data of the corresponding sub-model, the final third data, and the latent space information.
18. The defect prediction device of claim 11 , wherein the processor is configured to divide the production process data based on steps.
19. The defect prediction device of claim 11 , wherein the processor is configured to determine the sequence between the plurality of sub-models based on any one of or any combination of a temporal order and a physical correlation of the production process.