Neural networks for enhanced stress prognostics for encapsulated electronic packages - A comparison

Meszmer, Peter; Majd, Mehryar; Prisacaru, Alexandru; Gromala, Przemyslaw Jakub; Wunderle, Bernhard

doi:10.1016/j.microrel.2021.114181

Cited by 8 publications

(2 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Yuan et al [4] developed a gated neural network technique to learn the performance shifting of the solid-state lighting (SSL) lamp over time. Meszmer et al [5] applied many NN techniques to study which is the best for the electronic packaging, and the sequential NN performed best, including the gate recurrent unit (GRU) and long short-term memory (LSTM) architectures because of their capability to capture the characteristics of the sequential dataset. Selvanayagam et al [6] applied the AI-assisted modeling concept for the improvement of the packaging warpage.…”

Section: Introductionmentioning

confidence: 99%

Solder Joint Reliability Risk Estimation by AI-Assisted Simulation Framework with Genetic Algorithm to Optimize the Initial Parameters for AI Models

2021

View full text Add to dashboard Cite

Solder joint fatigue is one of the critical failure modes in ball-grid array packaging. Because the reliability test is time-consuming and geometrical/material nonlinearities are required for the physics-driven model, the AI-assisted simulation framework is developed to establish the risk estimation capability against the design and process parameters. Due to the time-dependent and nonlinear characteristics of the solder joint fatigue failure, this research follows the AI-assisted simulation framework and builds the non-sequential artificial neural network (ANN) and sequential recurrent neural network (RNN) architectures. Both are investigated to understand their capability of abstracting the time-dependent solder joint fatigue knowledge from the dataset. Moreover, this research applies the genetic algorithm (GA) optimization to decrease the influence of the initial guessings, including the weightings and bias of the neural network architectures. In this research, two GA optimizers are developed, including the “back-to-original” and “progressing” ones. Moreover, we apply the principal component analysis (PCA) to the GA optimization results to obtain the PCA gene. The prediction error of all neural network models is within 0.15% under GA optimized PCA gene. There is no clear statistical evidence that RNN is better than ANN in the wafer level chip-scaled packaging (WLCSP) solder joint reliability risk estimation when the GA optimizer is applied to minimize the impact of the initial AI model. Hence, a stable optimization with a broad design domain can be realized by an ANN model with a faster training speed than RNN, even though solder fatigue is a time-dependent mechanical behavior.

show abstract

Section: Introductionmentioning

confidence: 99%

Solder Joint Reliability Risk Estimation by AI-Assisted Simulation Framework with Genetic Algorithm to Optimize the Initial Parameters for AI Models

2021

View full text Add to dashboard Cite

show abstract

“…Among other subject areas, we note, for example, microelectronics, where image blurring caused by noise strongly affects the quality of prediction [18] and observations of the biosphere. In [19], the authors note that biometric data (in a broad sense, a kind of data observations of the biosphere) often have observation gaps, outliers and breaks.…”

mentioning

confidence: 99%

To the question of restoring symbol sequences encoding noisy periodic functions

Zhukova¹,

Ulyanov²

2021

Bus. Inform.

View full text Add to dashboard Cite

In business informatics, one of the research subjects is the analysis of data on processes in applied subject areas; here problems of qualitative analysis arise. Such problems arise, for example, in the qualitative study of log files of business processes, in the analysis and prediction of time series and other processes of a different nature. Quite often, to represent information about the processes under study, the methods of qualitative analysis use symbolic coding, which makes it possible to remove unnecessary detailing of numerical descriptions. The relevance of this study is due to the fact that when working with the raw data, researchers often face the presence of noise and distortions of the data, which significantly complicates the solution of the problems of qualitative analysis. When working with symbolic representations of the processes under study, which quite often have a periodic nature, we observe noise of deletion, insertion and replacement of symbols, which complicate the solution of the problem of revealing and analyzing the periodicity. This article deals with the problem of recovering periodic symbolic sequences obtained by coding from samples of continuous periodic functions and distorted by noise of insertion, replacement and deletion of symbols. Trigonometric functions are considered as a specific example of synthetic time series data. To encode trigonometric functions, alphabets of various cardinalities are used. The article presents an experimental study of the dependence of the quality characteristics of the method of period and a periodically repeating fragment recovery, previously proposed by the authors and improved in this study. For alphabets of different cardinalities at fixed sampling intervals, the fraction of sequences with a satisfactorily reconstructed period and the relative error in determining the period are given. The quality of reconstruction of a periodically repeating fragment is estimated by the edit distance from the reconstructed periodic sequence to the original sequence distorted by noise.

show abstract

Heterogeneous and hybrid integration system in display technology

Huang,

Long,

et al. 2024

Moore. More

View full text Add to dashboard Cite

The development tendency of “More than Display” is proposed for the display and semiconductor technologies, and the new-brand architecture of heterogeneous integration system in display (HiSID) is established in accordance with the demands of third generation Micro/Mini-LED devices. Many functional units (e.g., display units, storage units, sensing units, communication units and computing units) are integrated into one display main-board based on the semiconductor technology and electronic packaging. The advantages and details of miniaturization, intelligent, advanced integration, signal integrity with low latency performance, process compatibility and reliability are introduced. The interconnection requirements and design of the HiSID model with artificial intelligence are also summarized in this paper. It will provide technical guidance and references for the commercial application, core technology, and breakthrough direction of the HiSID module in display technology.

show abstract

Neural networks for enhanced stress prognostics for encapsulated electronic packages - A comparison

Cited by 8 publications

References 26 publications

Solder Joint Reliability Risk Estimation by AI-Assisted Simulation Framework with Genetic Algorithm to Optimize the Initial Parameters for AI Models

Solder Joint Reliability Risk Estimation by AI-Assisted Simulation Framework with Genetic Algorithm to Optimize the Initial Parameters for AI Models

To the question of restoring symbol sequences encoding noisy periodic functions

Heterogeneous and hybrid integration system in display technology

Contact Info

Product

Resources

About