Multi-objective optimization: methods and applications

Multi-objective optimization is concerned with finding solutions to a decision problem with multiple, normally conflicting objectives. This chapter focusses on multi-objective optimization problems that can be characterised within the paradigm of mathematical programming. Three modelling techniques that are well established in the literature are presented: Pareto set generation, goal programming and compromise programming. Each method is described, along with its strengths, weaknesses and areas of application. The underlying assumptions and philosophies of each method, nature of interaction of decision makers and nature of solutions produced is discussed and compared between the three methods. A small but representative example is given for each method and the results are discussed and conclusions are drawn.