No abstract

No abstract

This investigation presents a data-driven Long-short Term Memory battery model for predicting State of Charge for lithium-ion batteries LiFePO4 for next-generation vehicle operations. Our modified algorithm builds and updates a model using multivariate inputs that include physical properties, voltage, current, and ambient temperature during operations. The primary research goal is to improve prediction performance on future values from multiple training examples using an online learning scheme. Initial results demonstrate excellent predictions that outperform results from literature and other neural network algorithms. Due to computing constraints in on-board vehicle systems, the authors develop online training with autonomous control of lag (window width). The control algorithm embeds in the model with rules that govern and adjust lag during training. This method ensures the minimization of computational cost and prediction errors with the use of standard computing equipment during driving conditions.

Defects are a leading issue for the rejection of parts manufactured through the Directed Energy Deposition (DED) Additive Manufacturing (AM) process. In an attempt to illuminate and advance in situ quality monitoring and control of workpieces, we present an innovative data-driven method that synchronously collects sensing data and AM process parameters with a low sampling rate during the DED process. The proposed data-driven technique determines the important influences that individual printing parameters and sensing features have on prediction at the inter-layer qualification to perform feature selection. Three Machine Learning (ML) algorithms including Random Forest (RF), Support Vector Machine (SVM), and Convolutional Neural Network (CNN) are used. During post-production, a threshold is applied to detect low-density occurrences such as porosity sizes and quantities from CT scans that render individual layers acceptable or unacceptable. This information is fed to the ML models for training. Training/testing are completed offline on samples deemed “high-quality” and “low-quality”, utilizing only features recorded from the build process. CNN results show that the classification of acceptable/unacceptable layers can reach between 90% accuracy while training/testing on a “high-quality” sample and dip to 65% accuracy when trained/tested on “low-quality”/“high-quality” (respectively), indicating over-fitting but showing CNN as a promising inter-layer classifier.