Multi-source modeling with legacy data

Abstract

A method for estimating a crack propagation rate includes receiving a first dataset for a new material design, the new material design including crack growth rate data, receiving second and third datasets for a plurality of different legacy systems associated with an existing material design, determining a legacy model for each of the plurality of different legacy systems based on the respective second and third datasets for each of the plurality of different legacy systems and the first dataset for the new material design, the legacy model based on a validity of each legacy system, the validity determined using a legacy model likelihood validity and a predictive uncertainty model validity, calculating a model weight to associate with each of the determined legacy models, and determining a first multi-source model for new data for the new material design based on the model weight.

What is claimed is: 
     
         1 . A method for estimating a crack propagation rate, comprising:
 receiving a first dataset for a new material design, the new material design including crack growth rate data;   receiving second and third datasets for a plurality of different legacy systems associated with an existing material design;   determining a legacy model for each of the plurality of different legacy systems based on the respective second and third datasets for each of the plurality of different legacy systems and the first dataset for the new material design, the legacy model based on a validity of each legacy system, the validity determined using a legacy model likelihood validity and a predictive uncertainty model validity;   calculating a model weight to associate with each of the determined legacy models;   determining a first multi-source model for new data for the new material design based on the model weight; and   determining a predictive analysis for a new dataset for the new material design based on the first multi-source model, the predictive analysis including an estimated crack propagation rate for the new material design.   
     
     
         2 . The method of  claim 1 , wherein the first, second, or third dataset includes the crack growth rate data for at least one of a loading frequency, a stress ratio, or an amplitude stress intensity factor. 
     
     
         3 . The method of  claim 1 , wherein the legacy model is built as a function of at least one of a loading frequency, a stress ratio, or an amplitude stress intensity factor. 
     
     
         4 . The method of  claim 1 , further estimating the crack propagation rate based on a second multi-source model. 
     
     
         5 . The method of  claim 4 , wherein a model prediction accuracy of the second multi-source model is higher than the model prediction accuracy of the first multi-source model. 
     
     
         6 . The method of  claim 1 , wherein each legacy model is determined based on at least one of (1) a model for legacy data for each respective legacy model or (2) a discrepancy model indicative of a discrepancy from the legacy system to the new material design. 
     
     
         7 . The method of  claim 1 , wherein the first multi-source model is valid to accurately predict an outcome for all new data for the new material design, within a relevant range of input variables. 
     
     
         8 . An apparatus comprising:
 interface circuitry;   machine readable instructions; and   programmable circuitry to at least one of instantiate or execute the machine readable instructions to:   receive a first dataset for a new material design, the new material design including crack growth rate data;   receive second and third datasets for a plurality of different legacy systems associated with an existing material design;   determine a legacy model for each of the plurality of different legacy systems based on the respective second and third datasets for each of the plurality of different legacy systems and the first dataset for the new material design, the legacy model based on a validity of each legacy system, the validity determined using a legacy model likelihood validity and a predictive uncertainty model validity;   calculate a model weight to associate with each of the determined legacy models;   determine a first multi-source model for new data for the new material design based on the model weight; and   determine a predictive analysis for a new dataset for the new material design based on the first multi-source model, the predictive analysis including an estimated crack propagation rate for the new material design.   
     
     
         9 . The apparatus of  claim 8 , wherein the first, second, or third dataset includes the crack growth rate data for at least one of a loading frequency, a stress ratio, or an amplitude stress intensity factor. 
     
     
         10 . The apparatus of  claim 8 , wherein the programmable circuitry is to build the legacy model as a function of at least one of a loading frequency, a stress ratio, or an amplitude stress intensity factor. 
     
     
         11 . The apparatus of  claim 8 , wherein the programmable circuitry is to generate the estimated crack propagation rate based on a second multi-source model. 
     
     
         12 . The apparatus of  claim 11 , wherein a model prediction accuracy of the second multi-source model is higher than the model prediction accuracy of the first multi-source model. 
     
     
         13 . The apparatus of  claim 8 , wherein the programmable circuitry is to determine each legacy model based on at least one of (1) a model for legacy data for each respective legacy model or (2) a discrepancy model indicative of a discrepancy from the legacy system to the new material design. 
     
     
         14 . The apparatus of  claim 8 , wherein the first multi-source model is valid to accurately predict an outcome for all new data for the new material design, within a relevant range of input variables. 
     
     
         15 . A non-transitory machine readable storage medium comprising instructions to cause programmable circuitry to at least:
 receive a first dataset for a new material design, the new material design including crack growth rate data;   receive second and third datasets for a plurality of different legacy systems associated with an existing material design;   determine a legacy model for each of the plurality of different legacy systems based on the respective second and third datasets for each of the plurality of different legacy systems and the first dataset for the new material design, the legacy model based on a validity of each legacy system, the validity determined using a legacy model likelihood validity and a predictive uncertainty model validity;   calculate a model weight to associate with each of the determined legacy models;   determine a first multi-source model for new data for the new material design based on the model weight; and   determine a predictive analysis for a new dataset for the new material design based on the first multi-source model, the predictive analysis including an estimated crack propagation rate for the new material design.   
     
     
         16 . The non-transitory machine readable storage medium of  claim 15 , wherein the first, second, or third dataset includes the crack growth rate data for at least one of a loading frequency, a stress ratio, or an amplitude stress intensity factor. 
     
     
         17 . The non-transitory machine readable storage medium of  claim 15 , wherein the instructions are to cause the programmable circuitry to build the legacy model as a function of at least one of a loading frequency, a stress ratio, or an amplitude stress intensity factor. 
     
     
         18 . The non-transitory machine readable storage medium of  claim 15 , wherein the instructions are to cause the programmable circuitry to generate the estimated crack propagation rate based on a second multi-source model. 
     
     
         19 . The non-transitory machine readable storage medium of  claim 18 , wherein a model prediction accuracy of the second multi-source model is higher than the model prediction accuracy of the first multi-source model. 
     
     
         20 . The non-transitory machine readable storage medium of  claim 15 , wherein the instructions are to cause the programmable circuitry to determine each legacy model based on at least one of (1) a model for legacy data for each respective legacy model or (2) a discrepancy model indicative of a discrepancy from the legacy system to the new material design.