Kuntal Maji

Forward and reverse mappings of electrical discharge machining process using adaptive network-based fuzzy inference system

Input-output relationships of an electrical discharge machining process have been established both in forward as well as reverse directions using adaptive network-based fuzzy inference system. Three input parameters, such as peak current, pulse-on-time and pulse-duty-factor, and two outputs, namely material removal rate and surface roughness have been considered for the said mappings. A batch mode of training has been adopted with the help of 1000 data for the developed adaptive network-based fuzzy inference system, which has been designed using linear (say triangular) and non-linear (bell-shaped) membership function distributions of the input variables, separately. The performances of the developed models have been tested for both forward and reverse mappings with the help of some test cases collected through the real experiments. Adaptive network-based fuzzy inference system is found to tackle the problems of forward and reverse mappings efficiently. The fuzzy inference system utilizing non-linear membership functions is seen to perform slightly better than that with linear membership functions for the input variables.

Analysis and synthesis of laser forming process using neural networks and neuro-fuzzy inference system

To apply laser forming process in reality, it is required to know the relationships between the deformed shape and scanning paths along with heating conditions. The deformation due to laser scanning depends on various factors, namely laser power, scan speed, spot diameter, scan position, number of scans, and many others. This article presents soft computing-based methods to predict deformations for a set of heating conditions, and also to determine the heating lines and heat conditions, in order to get a desired shape (i.e., inverse analysis). A novel attempt has been made in this paper to carry out analysis and synthesis (inverse analysis) of laser forming process using both genetic-neural network (GA-NN) and genetic adaptive neuro-fuzzy inference system (GA-ANFIS). During the analysis, laser power, scan speed, spot diameter, scan position and number of scans are taken as inputs and bending angle is considered as the output. A batch mode of training has been used for both the approaches with the help of some experimental data. The performances of the developed approaches have been tested on some real experimental data. Both the approaches are found to be effective to predict the bending angles and carry out the process synthesis successfully. GA-NN approach is found to perform better than the GA-ANFIS approach in predicting the bending angles, and both the approaches are able to provide comparable predictions in inverse analysis.

Modelling of electrical discharge machining process using regression analysis, adaptive neuro-fuzzy inference system and genetic algorithm

Input-output relationships of an electrical discharge machining process have been determined based on some experimental data (collected according to a non-rotatable and face-centred central composite design) using statistical regression analysis and adaptive neuro-fuzzy inference system. Three input parameters, such as peak current, pulse-on-time and pulse-duty-factor and two outputs, namely material removal rate and surface roughness have been considered for the said modelling. The performances of the developed models have been checked using some test cases collected through the real experiments. Both single- as well as multi-objective optimisation problems have been formulated and solved using genetic algorithm. A set of optimal input parameters has been identified to ensure the maximum material removal rate and minimum surface roughness. An interesting Pareto-optimal front of solutions has also been obtained.

Analysis of pulsed laser bending of sheet metal using neural networks and neuro-fuzzy system

Laser bending is a nontraditional forming process, where sheet metal gets plastically deformed by laser-induced thermal stresses. The objective of this study is to establish the relationships between bending angles and process parameters in a pulsed laser bending process using soft computing–based methods, that is, neural networks and neuro-fuzzy system. Laser power, scan speed, spot diameter and pulse duration were considered as inputs, and bending angle was taken as output for modeling the bending angle (called forward analysis). In the case of inverse analysis or process synthesis (i.e. to determine the process parameters in order to achieve the desired outputs), bending angle and pulse duration were considered as inputs, and laser power, scan speed and spot diameter were treated as outputs. For both forward and inverse analyses, neural networks and neuro-fuzzy systems were trained in a batch mode with experimental data using two different algorithms, that is, genetic algorithm and back-propagation algorithm. The optimized networks were used for the predictions of bending angles and process parameters for some test cases. All the developed models were found to be satisfactory for both the analyses. Genetic algorithm was found to perform better than the back-propagation algorithm for both the networks in terms of prediction accuracy but at the cost of computational time. Neural networks trained with genetic algorithm were seen to perform better than the other models in predicting bending angles and process parameters. The developed models might be helpful in automating the pulsed laser forming process.

Numerical and Experimental Studies on Pulsed Laser Forming of Sheet Metal

Pulsed laser forming is a non-contact thermal forming process, where sheet metal gets plastically deformed by thermal residual stresses induced by controlled discontinuous laser irradiations. The temperature and deformation fields have been determined using finite element analysis under different processing conditions. Two types of pulsed laser forming processes, i.e., overlapped and discrete spot forming have been identified depending on the combinations of process parameters. Bending angle is found to increase with the degree of overlap and decrease with the increase of gap in case of the two types of spot forming processes. A comparative study between pulsed and continuous laser forming has also been performed using both finite element simulations and experiments. Bending angle in case of discrete spot pulsed laser forming is found to be more compared to the continuous laser forming. The results of finite element simulations have been found to be in good agreement with the experimental results.