A multi-objective algorithm for optimization of modern machining processes. (May 2017)

• Record Type: Journal Article
• Authors: Rao, R. Venkata; Rai, Dhiraj P.; Balic, Joze
• Is Part Of: Engineering applications of artificial intelligence. Volume 61(2017:Jan.)
• Page Start: 103
• Page End: 125
• Publication Date: 2017-05
• Subjects: Modern machining processes -- Process parameters -- Multi-objective optimization -- A posteriori approach -- Jaya algorithm -- Population based algorithms; Engineering -- Data processing -- Periodicals; Artificial intelligence -- Periodicals; Expert systems (Computer science) -- Periodicals; Ingénierie -- Informatique -- Périodiques; Intelligence artificielle -- Périodiques; Systèmes experts (Informatique) -- Périodiques; Artificial intelligence; Engineering -- Data processing; Expert systems (Computer science); Periodicals; 620.00285
• DOI: 10.1016/j.engappai.2017.03.001
• Languages: English
• ISSNs: 0952-1976

Abstract

Abstract: Multi-objective optimization aspects of four modern machining processes namely wire-electro discharge machining process, laser cutting process, electrochemical machining process and focused ion beam micro-milling process are considered in this work. In WEDM process cutting velocity and surface quality are important objectives which are mutually conflicting in nature. Minimization of kerf taper is vital in the laser cutting process which increases with the increase in material removal rate. The ECM process is characterized by high material removal rate, but poor dimensional accuracy, high tool wear rate and high over cut. FIB micro-milling process is useful in applications where a nano-level surface finish is desired but this process is characterized by a very low material removal rate. All the above mentioned objectives are vital as they closely govern the performance of the machining processes considered in this work. Therefore, the aim of this work is to achieve these objectives through process parameter optimization. In order to handle multiple objectives simultaneously a new posteriori multi-objective optimization algorithm named as multi-objective Jaya (MO-Jaya) algorithm is proposed which can provide multiple optimal solutions in a single simulation run. The regression models for the above mentioned machining processes which were developed by previous researchers are used as fitness function for MO-Jaya algorithm. In the case of WEDM process the optimization problem is an unconstrained, linear and parameter bounded. In the case of laser cutting process the optimization problem is a non-linear, unconstrained, quadratic and parameter bounded. In the ECM process the optimization problem is a non-linear, unconstrained, quadratic and parameter bounded. The second case study of ECM process the optimization problem is a non-linear, constrained, non-quadratic and parameter bounded. In the case of FIB micro-milling process, the optimization problem is a non-linear, unconstrained, quadratic and parameter bounded. In addition, the performance of MO-Jaya algorithm is also tested on a non-linear, non-quadratic unconstrained multi-objective benchmark function of CEC2009. In order to handle the constraints effectively a heuristic approach for handling constraints known as the constrained-dominance concept is used in MO-Jaya algorithm. In order to ensure that the newly generated solutions are within the parameter bounds a parameter-bounding strategy is used in MO-Jaya algorithm. The results of MO-Jaya algorithm are compared with the results of GA, NSGA, NSGA-II, BBO, NSTLBO, PSO, SQP and Monte Carlo simulations. The results have shown the better performance of the proposed algorithm. Graphical abstract: Flowchart for the MO-Jaya algorithm. Highlights: A new multi-objective optimization algorithm named as MO-Jaya algorithm is proposed. Multi-objective optimization aspects of modern machining processes are considered. A Pareto optimal set of solutions along with a Pareto front is obtained for each of the considered machining processes. The MO-Jaya algorithm may also be applied to multi-objective optimization problems of other manufacturing processes.