Reconstruction of the thermal conductivity coefficient by using the harmony search algorithm

Purpose: of this paper: Aim of this paper is a presentation of the respectively new tool for solving the optimization problems, which is the Harmony Search algorithm in version slightly modified by the authors, used for identifying the thermal conductivity coefficient. Proposed approach is illustrated with an example confirming its usefulness for solving such kinds of problems. Design/methodology/approach: For solving the considered parametric inverse heat conduction problem the approach is applied in which the essential part consists in minimization of the functional representing the differences between the measurement values of temperature and approximate values calculated with the aid of finite difference method. For minimizing the functional the Harmony Search algorithm is used. Findings: The elaboration shows that approaches involving the algorithms of artificial intelligence for solving the inverse heat conduction problems of that kind are efficient and they ensure to receive satisfying results in shorter time in comparison with the classical procedures. Research limitations/implications: Specific properties of the heuristic algorithms require to execute the procedure several times and to average the obtained results because each running of the algorithm can give slightly different results. Each execution of the procedure means the solution of the direct problem associated with the considered inverse problem by using the finite difference method. Practical implications: In spite of the problem described above the approaches involving the heuristic algorithms of artificial intelligence are successful because they are respectively simple and easy to use and they give satisfying results after short time of working. Another advantage of using optimization algorithms of that kind is the fact that they do not need to satisfy any assumptions about the solved problem, oppositely to the classical optimization algorithms. Originality/value: Proposal of the original approach involving the heuristic optimization algorithm for solving the parametric inverse heat conduction problem is discussed in the paper.