Approach to prediction of laser cutting quality by employing fuzzy expert system

Research highlights? Fuzzy expert system to predict the effect in CO2 laser cutting process was developed using MATLAB FL Toolbox. The relationship between laser cutting parameters (power, speed and gas pressure) and FL predicted cutting output characteristic (surface roughness and dross inclusion) was established. Appropriate selection parameters in fuzzy inference process leads to a good correlation between the experimental and prediction result. Development of the FL model provide a good foundation and encouragement for laser industrialist to expand the use of lasers in materials processing. In laser cutting, the surface quality and metallurgical characteristics of the end product is what matters the most in terms of laser cutting quality. Thus, an attempt has been carried out to develop an expert system using fuzzy logic model to predict the effect of carbon dioxide (CO2) laser cutting quality based on laser cutting parameters onto 1mm thickness of Incoloy? alloy 800. The predicting fuzzy logic model is implemented on Fuzzy Logic Toolbox of MATLAB using Mamdani technique. A set of training and testing consists of 125 data used in the fuzzy logic model are arranged in a format of three input parameters that cover the power, assist gas pressure and cutting speed, and two output parameters which are the surface roughness and dross inclusion. The relationships between experimental results, fuzzy logic model and statistical results for both training and testing performance exhibited a good correlation. Based on the results of the study, it shows that the proposed fuzzy logic model can be used to predict the surface roughness and dross inclusion of carbon dioxide (CO2) laser cutting process for Incoloy? alloy 800.