Systems, and methods for diagnosing an additive manufacturing device using a physics assisted machine learning model
A system for diagnosing an additive manufacturing device is provided. The system includes a first module configured to: obtain one or more parameters for a digital twin of a component of the additive manufacturing device based on raw data from the component of the additive manufacturing device; and generate physics features for the digital twin of the component of the additive manufacturing device based on the one or more parameters and one or more transfer functions, a second module configured to obtain one or more classifiers for classifying the component as a first condition or a second condition based on physics features; and a third module configured to: determine a health of the component based on the generated physics features of the first model and the one or more classifiers.
1. A system for diagnosing an additive manufacturing device, the system comprising: a processor; and memory, the memory including: a physics-based model forming a digital twin of at least a portion of the additive manufacturing device; a data science model formed of at least one of a trained machine learning model or a statistical model; and a diagnostic model to: receive a first input from the physics-based model and a second input from the data science model; determine, using the first input and the second input, a root cause of a build issue; and generate a correction for the build issue based on the determined root cause; the processor to implement an adjustment module to adjust one or more parameters of the additive manufacturing device to perform the correction in accordance with the adjusted one or more parameters.
2. The system of claim 1 , wherein the physics-based model includes one or more physics features, and wherein the data science model includes one or more classifiers.
3. The system of claim 2 , further including a hybrid classification model, the hybrid classification model to determine a state of at least the portion of the additive manufacturing device by comparing the one or more physics features with threshold values of the one or more classifiers.
4. The system of claim 3 , wherein the hybrid classification model uses a weighted average method with the first input and the second input to classify health of the at least a portion of the additive manufacturing device, the first input relating to a deviation from the physics-based model and the second input relating to a deviation from the data science model.
5. The system of claim 3 , wherein the hybrid classification model is to predict an amount of time till next action.
6. The system of claim 2 , wherein the diagnostic model is to determine the root cause by comparing the one or more physics features with the one or more classifiers.
7. The system of claim 2 , wherein the physics-based model includes one or more transfer functions.
8. The system of claim 7 , wherein the one or more transfer functions is updated based on parameters related to wear and tear of the at least a portion of the additive manufacturing device.
9. The system of claim 2 , wherein the data science model is to update the one or more classifiers based on parameters related to wear and tear of the at least a portion of the additive manufacturing device.
10. The system of claim 1 , further including a feature extraction module to extract features from data and provide the features to at least one of the physics-based model or the data science model.
11. A non-transitory machine readable medium including computer executable instructions that, when executed by one or more processors, cause the one or more processors to implement at least: a physics-based model forming a digital twin of at least a portion of an additive manufacturing device; a data science model formed of at least one of a trained machine learning model or a statistical model; and a diagnostic model to receive a first input from the physics-based model and a second input from the data science model; determine, using the first input and the second input, a root cause of a build issue; and generate a correction for the build issue based on the determined root cause;
the one or more processors to adjust one or more parameters of the additive manufacturing device and perform the correction in accordance with the adjusted one or more parameters.
12. The non-transitory machine readable medium of claim 11 , wherein the physics-based model includes one or more physics features, and wherein the data science model includes one or more classifiers.
13. The non-transitory machine readable medium of claim 12 , further including a hybrid classification model, the hybrid classification model to determine a state of at least the portion of the additive manufacturing device by comparing the one or more physics features with threshold values of the one or more classifiers.
14. The non-transitory machine readable medium of claim 13 , wherein the hybrid classification model uses a weighted average method with the first input and the second input to classify health of the at least a portion of the additive manufacturing device, the first input relating to a deviation from the physics-based model and the second input relating to a deviation from the data science model.
15. The non-transitory machine readable medium of claim 13 , wherein the hybrid classification model is to predict an amount of time till next action.
16. The non-transitory machine readable medium of claim 12 , wherein the diagnostic model is to determine the root cause by comparing the one or more physics features with the one or more classifiers.
17. The non-transitory machine readable medium of claim 12 , wherein the physics-based model includes one or more transfer functions.
18. The non-transitory machine readable medium of claim 17 , wherein the one or more transfer functions is updated based on parameters related to wear and tear of the at least a portion of the additive manufacturing device.
19. The non-transitory machine readable medium of claim 12 , wherein the data science model is to update the one or more classifiers based on parameters related to wear and tear of the at least a portion of the additive manufacturing device.
20. A method comprising: generating a digital twin of at least a portion of an additive manufacturing device using a physics-based model; generating a data science model using at least one of a trained machine learning model or a statistical model; receiving, using a diagnostic model, a first input from the physics-based model and a second input from the data science model; determining, by the diagnostic model using the first input and the second input, a root cause of a build issue; generating, using the diagnostic model, a correction for the build issue;
and adjusting one or more parameters of the additive manufacturing device to perform the correction in accordance with the adjusted one or more parameters.