Comparison of ANN and DoE for the prediction of laser-machined micro-channel dimensions

Abstract

This paper presents four models developed for the prediction of the width and depth dimensions of CO 2 laser formed micro-channels in glass. A 3 3 statistical design of experiments (DoE) model was built and conducted with the power (P), pulse repetition frequency (PRF), and traverse speed (U) of the laser machine as the selected parameters for investigation. Three feed-forward, back-propagation Artificial Neural Networks (ANNs) models were also generated. These ANN models were varied to investigate the influence of variations in the number and the selection of training data. Model A was constructed with 24 data randomly selected from the experimental results, leaving three data points for model testing; Model B was constructed with the eight corner points of the experimental data space, and seven other randomly selected data, leaving 12 data points for testing; and Model C was constructed with 15 randomly selected data leaving 12 data points for testing.These models were developed separately for both micro-channel width and depth prediction. These ANN models were constructed in LabVIEW coding. The performance of these ANN models and the DoE model were compared. When compared with the actual results two of the ANN models showed greater average percentage error than the DoE model. The other ANN model showed an improved predictive capability that was approximately twice as good as that provided from the DoE model.

Modeling pulsed laser micromachining of micro geometries using machine-learning techniques

“…Kumar et al (Kumar, 2010) investigated the influence of laser power, pulse frequency, number of scans and air pressure, on the groove depth in the generation of micro-notches with a nanosecond pulsed fiber laser on stainless steel and aluminum. Karazi et al (Karazi, 2009) machined and characterized micro-channel formation by laser machining. They studied the effects of laser power, pulse frequency and scanning speed on the width and depth of the channels.…”

“…Most of the very few works on this topic focus on the application of ANNs to this task: the work of Desai et al (Desai, 2012) predicted the depth of cut for singlepass laser micro-milling process using ANN and genetic programming approaches and the work of Karazi et al (Karazi, 2009) compared ANN and DoE models for the prediction of lasermachined micro-channel dimensions. If we open the state of the art to the application of AI techniques to machining processes similar to laser milling, we can conclude that ANNs are the most common technique used for most of these processes such as milling, drilling or laser finishing (Chandrasekaran, 2010), although many other AI techniques have also been applied for such purposes.…”

Laser machined macro and micro structures on glass for enhanced light trapping in solar cells

“…Duty cycle (%), pulse repetition frequency (Hz), and traverse speed (mm/s) were the main parameters considered in the investigation. After initial screening experiments were completed, the maximum and minimum values of each parameter were fixed and a 3 3 general factorial design of experiments (DOE) was completed with the value indicated in Table 1 [16]. The textures were created by raster scanning the laser beam over the glass sample to produce a texture of parallel grooves measuring eight by eight millimetres.…”

Evaluation of the Effect of Nd:yvo4 Laser Parameters on Internal Micro-Channel Fabrication in Polycarbonate

“…Various statistical and numerical methodologies have been implemented to predict and optimise several laser manufacturing processes including Artificial Neural Networks (ANN) (Lee et al 2001); Genetic Algorithms (GA) (Ye, Yuan and Zhou, 2009), Design of Experiments (DoE) (Karazi, Issa and Brabazon, 2009), Finite Element Analysis (FEA) (de Deus and Mazumder, 1996), Ant Colony optimisation (AC) (Wang and Xie, 2005), and Fuzzy Logic (FL) (Shen et al 2006). …”