Glazing Thickness Deposition Rate Modeling And Prediction For Robotic Glazing Of Hygiene Ceramic Ceramics

It is the basic and critical conditions of realizing automatic track-planning of robot offline programming tasks to understand the formation of glaze deposition scientifically and to establish the glaze thickness deposition rate model. By a series of experiments and optimization methods, a point on the body glaze deposition model is established. Model can plan Spray enamel process and gun trajectory, achieve specified thickness and precision and provide a basis of theory and programming. This thesis has a further research on the influence of material deposition patterns from tool orientation and material mobility rate and build corresponding glaze deposition pattern model.Based on the situation of spray enamel deposition rate from air spray gun, flat work piece spray enamel is tested. Glaze thickness deposition rate model is constructed by using artificial neural networks Bayesian normalization algorithm and LM optimization algorithm respectively on the basis of the test data of the glaze film thickness deposition rate. In MATLAB-based environment, the glazing deposition rate model is fitted by real-coded genetic algorithm to construct the glaze thickness deposition rate model with specific expression.By comparative analysis between simulation and experimental data, neural networks model of the glaze thickness deposition rate constructed by Bayesian normalization algorithm and LM optimization algorithm coincide with the measured data basically. The actual data of complex nonlinear glaze thickness deposition rate model are fitted very well and also there is high accuracy. It is verified that the error of thickness model is within 5μm, which tells that the glaze thickness deposition rate model is fitted properly and effectively. After further comparative analysis, better Bayesian algorithm deposition model is selected. By mixed programming with MATLAB and VC, its new fitting software is developed including genetic algorithm model fitting guide and producing model report form. The comparison of predicted and actual enamel thickness shows that the developed system is effective and meets the engineering.