Optimization System for Solving Problems of Nonlinear Multiobjective Programming

Every day almost everyone has to plan or organize some activity by using different resources time, money, etc. It is even more important for managers that often are in a continuous process of decision making. Most managers rely mainly on their experience and intuition, rather than on strict mathematical models and optimization methods. With advances in the computer industry, it is becoming far easier for these processes to be aided by software optimization systems. Development of different types of computer programs eases and helps the decision making process. A good example of such systems is MONP-16, designed in the Institute of Information Technologies, Bulgarian Academy of Sciences. It solves the basic class of multiobjective nonlinear programming problems. It runs on personal computers IBM PC/XT/AT and their compatibles and MS-DOS operating system. This system is designed basically for experts in mathematical programming as it requires a precise analysis and description of the problem as a multiobjective nonlinear programming one by the user. MOLP-16 uses the Satisfactory Trade-off method of Nakayama and single objective optimization Lagrange method (using Lagrange multipliers) as wells as SQP (sequential quadratic programming) method [2]. Computers are becoming faster and faster and operation systems easier and more user friendly. The appearance and dramatic growth of the global Internet with its graphical user interface lead up to the development and usage of modern technologies and new types of optimization systems for decision making. The Finland WWW-NIMBUS [3, 4], designed in the University of Jyvaskyla is an example of such a system. The NIMBUS algorithm is capable of solving nonlinear and real-world applications involving nondifferentiable and nonconvex functions. It was first implemented in Fortran for MS-DOS operating system and it was mainly БЪЛГАРСКА АКАДЕМИЯ НА НАУКИТЕ . BULGARIAN ACADEMY OF SCIENCES