Judging method for a module peeling time of a soft electronic fabric module and a system applying the same
A judging method for a module peeling time of a soft electronic fabric module is provided. The method includes: preselecting a plurality of module material combinations, the plurality of module material combinations respectively comprising a substrate material, a wire material and a packaging material; extracting the plurality of module material combinations to generate a plurality of module material combination parameters; generating a plurality of machine learning training data based on the plurality of module material combination parameters and a plurality of module pre-processing conditions; and training a machine learning model according to the plurality of machine learning training data to provide an optimized prediction model for judging a module peeling time.
1 . A judging method for a module peeling time of a soft electronic fabric module, comprising:
preselecting a plurality of module material combinations, the plurality of module material combinations respectively comprising a substrate material, a wire material and a packaging material;
extracting the plurality of module material combinations to generate a plurality of module material combination parameters;
generating a plurality of machine learning training data based on the plurality of module material combination parameters and a plurality of module pre-processing conditions; and
training a machine learning model according to the plurality of machine learning training data to provide an optimized prediction model for judging a module peeling time.
2 . The judging method for a module peeling time of a soft electronic fabric module according to claim 1 , wherein one of the plurality of module pre-processing conditions further comprises:
a process temperature range of 120° C. to 150° C. and a process time range of 13 minutes to 15 minutes.
3 . The judging method for a module peeling time of a soft electronic fabric module according to claim 1 , wherein one of the plurality of module material combination parameters further comprises:
a thermal expansion coefficient of the substrate material, a thermal expansion coefficient of the wire material, a thermal expansion coefficient of the packaging material, a heat resistance of the wire material, a heat resistance of the packaging material, and a heat resistance of the substrate material.
4 . The judging method for a module peeling time of a soft electronic fabric module according to claim 1 , wherein the machine learning model is a Bayesian algorithm.
5 . The judging method for a module peeling time of a soft electronic fabric module according to claim 4 , wherein the Bayesian algorithm further comprises a Scikit learn operation and a Gaussian Regression operation.
6 . A judging method for a module peeling time of a soft electronic fabric module, comprising:
preselecting a plurality of module material combinations, the plurality of module material combinations respectively comprising a substrate material, a wire material and a packaging material;
extracting the plurality of module material combinations to generate a plurality of module material combination parameters;
generating a plurality of machine learning training data based on the plurality of module material combination parameters and a plurality of module pre-processing conditions; and
training a machine learning model according to the plurality of machine learning training data to provide an optimized prediction model for judging a module peeling time,
wherein a heat resistance of the substrate material is 155° C. to 200° C., and a thermal expansion coefficient of the substrate material is 7 10 −6 /K to 147 10 −6 /K; a heat resistance of the wire material is 155° C. to 300° C., and a thermal expansion coefficient of the wire material is 117 10 −6 /K to 264 10 −6 /K; a heat resistance of the packaging material is 100° C. to 200° C., and a thermal expansion coefficient is 16 10 −6 /K to 380 10 −6 /K.
7 . The judging method for a module peeling time of a soft electronic fabric module according to claim 6 , wherein one of the plurality of module pre-processing conditions further comprises:
a process temperature range of 120° C. to 150° C. and a process time range of 13 minutes to 15 minutes.
8 . The judging method for a module peeling time of a soft electronic fabric module according to claim 6 , wherein one of the plurality of module material combination parameters further comprises:
a thermal expansion coefficient of the substrate material, a thermal expansion coefficient of the wire material, a thermal expansion coefficient of the packaging material, a heat resistance of the wire material, a heat resistance of the packaging material, and a heat resistance of the substrate material.
9 . The judging method for a module peeling time of a soft electronic fabric module according to claim 6 , wherein the machine learning model is a Bayesian algorithm.
10 . The judging method for a module peeling time of a soft electronic fabric module according to claim 9 , wherein the Bayesian algorithm further comprises a Scikit learn operation and a Gaussian Regression operation.
11 . A prediction system for a module peeling time of a soft electronic fabric module, comprising:
a data collection system, preselecting a plurality of module material combinations, wherein the plurality of module material combinations respectively comprises a substrate material, a wire material and a packaging material;
a measuring device, extracting the plurality of module material combinations to generate a plurality of module material combination parameters;
a lamination device, performing a plurality of module pre-processing conditions to the plurality of module material combinations;
an environmental testing device, measuring a plurality of module peeling time for training of the plurality of module material combinations after performing the plurality of module pre-processing conditions; and
a model building system, generating a plurality of machine learning training data based on the plurality of module material combination parameters and the plurality of module peeling time for training, and training a machine learning model according to the plurality of machine learning training data to provide an optimized prediction model for judging a module peeling time.
12 . The prediction system for a module peeling time of a soft electronic fabric module according to claim 11 , wherein one of the plurality of module pre-processing conditions further comprises:
a process temperature range of 120° C. to 150° C. and a process time range of 13 minutes to 15 minutes.
13 . The prediction system for a module peeling time of a soft electronic fabric module according to claim 11 , wherein one of the plurality of module material combination parameters further comprises:
a thermal expansion coefficient of the substrate material, a thermal expansion coefficient of the wire material, a thermal expansion coefficient of the packaging material, a heat resistance of the wire material, a heat resistance of the packaging material, and a heat resistance of the substrate material.
14 . The prediction system for a module peeling time of a soft electronic fabric module according to claim 11 , wherein the machine learning model is a Bayesian algorithm.
15 . The prediction system for a module peeling time of a soft electronic fabric module according to claim 14 , wherein the Bayesian algorithm further comprises a Scikit learn operation and a Gaussian Regression operation.