Bi-objective optimization of induction machine using interval-based interactive algorithms

Abstract

Purpose – Discrete highly constrained optimization of induction machine taking into consideration two objective functions: efficiency and total costs of production. The paper aims to discuss these issues. Design/methodology/approach – Interactive and semi-interactive interval-based optimization methods were used. Two concepts of multi-objective discrete optimization were proposed. Findings – Proposed methodology and algorithms allow decision maker (DM) participate in the process of optimal design and therefore decrease the total time of optimization process. The search procedure is straightforward and it does not require special skills of DM. Presented methods were successfully versified for the problem of optimal design with discrete variables. Research limitations/implications – Three interval algorithms suitable for inverse problems are researched and verified. It generally can be used for multi-objective problems. The dominance principles for interval boxes are showed in the paper. Proposed algorithms are based on the idea of hybridization of exact and evolutionary methods. Practical implications – Proposed approaches were successfully implemented within computer-aided application which is used by manufacturer of high power induction machine. Originality/value – The concept of pareto-domination using the interval boxes can be treated as original. The paper researched several elimination rules and discusses the difference between different approaches.